Record of DecisionRecord of Decision — Lower Duwamish Waterway Superfund Site iii consumption from...

Record of Decision

Lower Duwamish Waterway

Superfund Site

United States

Environmental Protection Agency

Region 10

November 2014

Record of Decision — Lower Duwamish Waterway Superfund Site

This blank page is for double sided printing.


i

Part 1 Declaration

Site Name and Location

Site Name: Lower Duwamish Waterway

Location: Seattle and Tukwila, King County, Washington

U.S. Environmental Protection Agency (EPA) identification number: WA00002329803

Statement of Basis and Purpose

This decision document presents the Selected Remedy for the In-waterway Portion of the Lower Duwamish

Waterway Superfund Site, in King County, Washington. The Selected Remedy was chosen in accordance

with the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA), and to the

extent practicable, the National Contingency Plan (NCP). This decision is based on the Administrative

Record file for this site. The State of Washington, through the Washington Department of Ecology, concurs

with the Selected Remedy.

Assessment of the Site

The response action selected in this Record of Decision (ROD) is necessary to protect the public health or

welfare or the environment from actual or threatened releases of hazardous substances into the environment.

Such a release or threat of release may present an imminent and substantial endangerment to public health,

welfare, or the environment.

Description of the Selected Remedy

The Selected Remedy is a final action for the In-waterway Portion of the Lower Duwamish Waterway

(LDW) Site. It addresses unacceptable human health risks associated with consumption of resident fish and

shellfish, and with direct contact (skin contact and incidental ingestion) from net fishing, clamming, and

beach play. It also addresses ecological risks to bottom-dwelling organisms (benthic invertebrates), fish, and

wildlife.

The Selected Remedy is the third component of an overall strategy for addressing contamination and the

associated risks in the LDW Site that includes:

1. early identification and cleanup of the most contaminated areas in the waterway, referred to as Early

Action Areas (EAAs) — an estimated 29 acres will be cleaned up in the EAAs;

2. controlling sources of contamination to the waterway (Washington State Department of Ecology

[Ecology] is the lead agency for this component); and

3. cleanup of the remaining contamination in the waterway, including long-term monitoring to assess the

success of the remedy in achieving cleanup goals (the Selected Remedy).

The Selected Remedy will be implemented after cleanup in the EAAs has been completed, source control

sufficient to minimize recontamination (see Section 4.2) has been implemented, additional sampling and

analysis has been conducted, and design of the remedy has been completed.


ii

The Selected Remedy addresses approximately 412 acres, and includes the following elements:

A total of 177 acres of active cleanup, consisting of:

– 105 acres of dredging or partial dredging and capping (an anticipated total volume of 960,000 cubic

yards would be dredged and disposed in an upland landfill);

– 24 acres of capping, with possible amendment with activated carbon or other contaminant-

sequestering agents; and

– 48 acres of Enhanced Natural Recovery (ENR – placing 6 to 9 inches of clean material over

contaminated sediments) with possible amendment with activated carbon or other contaminant-

sequestering agents, if these amendments are shown to be effective in pilot tests.

Further reduction of contaminant concentrations over time in the remaining 235 acres through

Monitored Natural Recovery (MNR – relying on natural processes such as burial of contaminated

sediments by cleaner sediments from upstream). Long-term monitoring data will determine whether

additional cleanup actions will be necessary in MNR areas.

– In MNR areas, more intensive long-term monitoring will be conducted in an estimated 33 acres

where contaminant of concern (COC) concentrations in sediment are less than the sediment remedial

action levels (RALs – contaminant concentrations above which remedial action is required) but

greater than the sediment cleanup objectives for protection of benthic invertebrates (benthic SCO);

this is referred to as MNR To Benthic SCO. If MNR does not achieve the benthic SCO or progress

sufficiently toward achieving it in 10 years, additional cleanup will be required as a part of this

remedy.

– Less intensive monitoring will be conducted in areas where sediment COC concentrations are below

the benthic SCO but above the sediment cleanup levels1 for protection of human health; this is

referred to as MNR Below Benthic SCO. This includes 202 acres where COC concentrations were

below the benthic SCO in remedial investigation sampling, and will also include the 33 acres

described in the previous bullet after COC concentrations are reduced to below the benthic SCO in

those areas. If the cleanup levels for protection of human health are not achieved, additional cleanup

actions will be considered in a future decision document.

Institutional controls (ICs) and LDW-wide monitoring, including:

– Proprietary controls, e.g., under the Washington Uniform Environmental Covenants Act (UECA), to

prohibit activity that could result in a release or exposure of COCs remaining in the subsurface

absent EPA approval; and

– Seafood consumption advisories.

The purpose of ICs is to protect the integrity of other remedial action elements such as capping, and to

provide information about how much and what types of fish and shellfish are safe to consume in the form of

fish advisories, education and outreach programs. A study is currently underway to gather information from

people who harvest or consume seafood and who may assist in understanding aspects of seafood

1 Cleanup levels are contaminant concentrations that must be achieved at the end of the 10-year natural recovery period.

They include human health-based levels (which must be met on an area-wide basis) and benthic SCO criteria (which

must be met on a point-by-point basis. See Section 8).


iii

consumption from the LDW as a first step in developing effective and appropriate ICs intended to reduce

exposure of the LDW seafood consuming community to risks from consuming resident fish and shellfish.

The Selected Remedy assumes completion of an additional 29 acres of cleanup in EAAs (see Section 4.1 for

further discussion of the EAAs), not included in the 412 acres addressed by the remedy.

The Selected Remedy is estimated to take 7 years to construct. The lowest contaminant concentrations in fish

and shellfish tissue are predicted by modeling to be achieved in 17 years following the start of construction.

Total estimated net present value costs (discounted at 2.3%) for the Selected Remedy are $342 million, of

which capital costs are $295 million, and operation, maintenance, and monitoring (OM&M) costs are

approximately $48 million.

Statutory Determinations

The Selected Remedy is protective of human health and the environment, complies with federal and state

requirements that are applicable or relevant and appropriate to the remedial action, is cost effective, and

utilizes permanent solutions and alternative treatment technologies to the maximum extent practicable for

this Site.

This remedy does not satisfy the statutory preference for treatment as a principal element of the remedy. The

NCP emphasizes the expectation that treatment will be used to address the principal threats posed by a site

whenever practicable. Principal threat waste is defined in EPA guidance as source material that is highly

toxic or highly mobile, and that generally cannot be contained in a reliable manner. EPA has determined that

the contaminated sediments in the LDW outside of the EAAs are not highly mobile or highly toxic. The

remedy does include potential treatment of some contaminated sediments through provisions for amendment

of caps and ENR with activated carbon or other contaminant-sequestering agents.

Because this remedial action will result in hazardous substances, pollutants, or contaminants remaining on-

site at levels above those that would allow for unlimited use and unrestricted exposure, statutory five-year

reviews will be conducted every five years after initiation of remedial action to ensure that the remedy

continues to be protective of human health and the environment.

ROD Data Certification Checklist

The following information is included in the Decision Summary (Part 2) of this ROD. Additional

information can be found in the Administrative Record for the site.

Contaminants of concern and their respective concentrations (Section 5.3)

Baseline risks represented by the contaminants of concern (Section 7).

Cleanup levels established for contaminants of concern and the basis for these levels (Section 8.2.1).

How source materials constituting principal threats are addressed (Sections 11 and 14.5).

Current and reasonably anticipated future land use assumptions and current and potential future

beneficial uses of surface water used in the baseline risk assessment and the ROD (Section 6).

Potential land and surface water use that will be available at the site as a result of the selected

remedy (Section 13.4).


v

Contents

Part 1 Declaration ......................................................................................................................... i

Acronyms and Abbreviations ...................................................................................................... xi

Part 2 Decision Summary ............................................................................................................ 1

1 Site Name, Location, and Brief Description ...................................................................... 1

2 Site History and Enforcement Activities ............................................................................ 2

2.1 Site History and Sources of Contamination ................................................................. 2

2.2 Previous Investigations ................................................................................................. 4

2.3 Cleanup Activities Planned and Completed to Date .................................................... 5

2.4 Source Control Investigations and Actions Completed to Date ................................... 7

2.5 Enforcement Activities ................................................................................................. 7

3 Community and Tribal Participation................................................................................. 9

4 Scope and Role of the Response Action............................................................................ 11

4.1 Component 1: Early Identification and Cleanup of EAAs ......................................... 11

4.2 Component 2: Controlling Sources of Contamination ............................................... 11

4.3 Component 3: In-Waterway Cleanup ......................................................................... 13

5 Site Characteristics ............................................................................................................ 15

5.1 Physical Characteristics .............................................................................................. 15

5.1.1 Surface Water Hydrology ...................................................................................... 15

5.1.2 Sediment Transport Model ..................................................................................... 16

5.1.3 Bed Composition Model ........................................................................................ 17

5.1.4 STM and BCM Uncertainty and Sensitivity Analyses........................................... 19

5.2 Contaminant Transfer Conceptual Site Model ........................................................... 19

5.2.1 Food Web Model ................................................................................................... 21

5.2.2 Food Web Model Uncertainty ................................................................................ 21

5.3 Nature and Extent of Contamination .......................................................................... 22

5.3.1 Surface and Subsurface Sediments ........................................................................ 22

5.3.2 Fish, Shellfish, and Benthic Invertebrate Tissue .................................................... 28

5.3.3 Surface Water ......................................................................................................... 29

5.3.4 Background COC Concentrations .......................................................................... 30

5.3.5 Sediment COC Concentrations from Upstream of the LDW Study Area .............. 32

6 Current and Potential Future Land and Waterway Use ................................................ 33

6.1 Land Use .................................................................................................................... 33

6.2 Waterway Use ............................................................................................................ 33


vi

7 Summary of Site Risks ....................................................................................................... 37

7.1 Human Health Risks ................................................................................................... 37

7.1.1 Identification of Contaminants of Potential Concern ............................................. 39

7.1.2 Exposure Assessment ............................................................................................. 40

7.1.3 Toxicity Assessment .............................................................................................. 47

7.1.4 Risk Characterization ............................................................................................. 48

7.1.5 Uncertainty Analysis for the HHRA ...................................................................... 57

7.2 Ecological Risks ......................................................................................................... 58

7.2.1 Ecological Communities in the LDW .................................................................... 58

7.2.2 Problem Formulation ............................................................................................. 59

7.2.3 Identification of Contaminants of Potential Concern for Ecological Receptors .... 59

7.2.4 Exposure and Effects Assessment .......................................................................... 59

7.2.5 Risk Characterization ............................................................................................. 64

7.2.6 Identification of COCs ........................................................................................... 65

7.2.7 Uncertainty Analysis for the ERA ......................................................................... 65

7.3 Basis for Action .......................................................................................................... 72

8 Remedial Action Objectives .............................................................................................. 73

8.1 Remedial Action Objectives ....................................................................................... 73

8.2 Cleanup Levels, ARARs and Target Tissue Concentrations ...................................... 73

8.2.1 Cleanup Levels ....................................................................................................... 74

8.2.2 ARARs ................................................................................................................... 76

8.2.3 Fish and Shellfish Target Tissue Concentrations ................................................... 78

9 Description of Alternatives ................................................................................................ 80

9.1 Framework for Developing Alternatives .................................................................... 80

9.2 Summary of Remedial Alternatives ........................................................................... 81

9.3 Technologies Common to All Remedial Alternatives ................................................ 87

9.4 Remedial Alternatives ................................................................................................ 89

10 Summary of Comparative Analysis of Alternatives ..................................................... 100

10.1 Threshold Criteria..................................................................................................... 101

10.1.1 Overall Protection of Human Health and the Environment ................................. 101

10.1.2 Compliance with ARARs ..................................................................................... 105

10.2 Balancing Criteria..................................................................................................... 109

10.2.1 Long-Term Effectiveness and Permanence ......................................................... 109

10.2.2 Reduction of Toxicity, Mobility, or Volume through Treatment ........................ 110

10.2.3 Short-Term Effectiveness .................................................................................... 111

10.2.4 Implementability .................................................................................................. 112


vii

10.2.5 Cost ...................................................................................................................... 112

10.3 Modifying Criteria .................................................................................................... 113

10.3.1 Community Acceptance ....................................................................................... 113

10.3.2 State/Tribal Acceptance ....................................................................................... 113

10.3.3 Environmental Justice Analysis ........................................................................... 113

10.4 Summary of CERCLA Nine-Criteria Evaluation ..................................................... 114

11 Principal Threat Waste ................................................................................................... 115

12 Documentation of Significant Changes to the Selected Remedy .................................. 116

13 Selected Remedy ............................................................................................................... 119

13.1 Summary of the Rationale for the Selected Remedy ................................................ 119

13.2 Description of the Selected Remedy ........................................................................ 120

13.2.1 Application of Cleanup Technologies .................................................................. 121

13.2.2 Monitored Natural Recovery ............................................................................... 128

13.2.3 Monitoring ........................................................................................................... 128

13.2.4 Institutional Controls............................................................................................ 130

13.2.5 Use of Green Remediation Practices.................................................................... 131

13.2.6 Role of EAAs in the Selected Remedy ................................................................ 131

13.2.7 Role of Source Control in the Selected Remedy .................................................. 131

13.2.8 Addressing Environmental Justice concerns ........................................................ 132

13.3 Cost Estimate for the Selected Remedy ................................................................... 132

13.4 Estimated Outcomes of Selected Remedy ................................................................ 134

14 Statutory Determinations ................................................................................................ 147

14.1 Protection of Human Health and the Environment .................................................. 147

14.2 Compliance with Applicable or Relevant and Appropriate Requirements .............. 147

14.3 Cost-Effectiveness .................................................................................................... 149

14.4 Utilization of Permanent Solutions and Alternative Treatment (or Resource Recovery)

Technologies to the Maximum Extent Practicable ................................................... 150

14.5 Preference for Treatment as a Principal Element ..................................................... 150

14.6 Five-Year Review Requirement ............................................................................... 150

15 Key Terms ......................................................................................................................... 151

16 References Cited ............................................................................................................... 155

Part 3 Responsiveness Summary [separate volume]............................................................... 161

Attachment 1 Washington State Department of Ecology Concurrence Letter ................... 163


viii

List of Tables

Table 1. Statistical Summaries for Baseline Human Health COC Concentrations in Sediment ..................... 23

Table 2. Summary of Selected Baseline Human Health COC Concentrations in Fish and Shellfish Tissuea 29

Table 3. Summary of PCB, Arsenic, cPAH, and Dioxin/Furan Data for Natural Background

Concentrations in Sediment ........................................................................................................... 30

Table 4. Summary of PCB, Arsenic, cPAH, and Dioxin/Furan Data for Natural Background

Concentrations in Fish and Shellfish Tissue .................................................................................. 31

Table 5. Estimates of Upstream Suspended Sediment Concentrations of PCBs, Arsenic, cPAHs, and

Dioxins/Furans Used in the LDW Bed Composition Model ......................................................... 32

Table 6. Rationale for Selection or Exclusion of Exposure Pathways in the Human Health Risk

Assessment .................................................................................................................................... 40

Table 7. Exposure Point Concentrations (EPCs) for Contaminants of Concern in Fish and Shellfish Used

in Human Health Risk Assessment ................................................................................................ 42

Table 8. Summary of Seafood Ingestion Exposure Parameters for Different Exposure Scenarios ................ 44

Table 9. Summary of Exposure Parameters for Direct Contact with Sediment for Different Exposure

Scenarios ........................................................................................................................................ 45

Table 10. Summary of Sediment Exposure Point Concentrations (EPCs) for Contaminants of Concern

Used in Human Health Risk Assessment and Updated in the Feasibility Study ........................... 45

Table 11. Cancer Toxicity Data Summary for Human Health COCs ............................................................. 49

Table 12. Noncancer Toxicity Data Summary for Human Health .................................................................. 50

Table 13. Summary of Cancer and Noncancer Risk Estimates for Human Health Scenarios ........................ 51

Table 14. Summary of COPCs and Rationale for Selection as COCs for Human Health Exposure

Scenarios ........................................................................................................................................ 56

Table 15. Numerical Benthic Sediment Cleanup Objectives and Benthic Cleanup Screening Levels from

the Washington State Sediment Management Standardsa ............................................................. 61

Table 16. Assessment Endpoints for Receptors of Concern (ROCs) and Measures of Effect and Exposure . 62

Table 17. Surface Sediment Contaminant Concentrations from FS Dataset, with Comparison to SMS

Chemical Criteria for Protection of Benthic Invertebrates ............................................................ 67

Table 18. Rationale for Selection of Contaminants as COCs for Ecological Risk ......................................... 69

Table 19. Cleanup Levels for PCBs, Arsenic, cPAHs, and Dioxins/Furans in Sediment for Human Health

and Ecological COCs (RAOs 1, 2 and 4) ...................................................................................... 74

Table 20. Sediment Cleanup Levels for Ecological (Benthic Invertebrate) COCs for RAO 3a ...................... 75

Table 21. LDW Resident Fish and Shellfish Target Tissue Concentrations ................................................... 79

Table 22. Remedial Alternatives and Associated Remedial Technologies, Remedial Action Levels, and

Actively Remediated Acres ........................................................................................................... 83

Table 23. Criteria for Assigning Recovery Categoriesa .................................................................................. 83

Table 24. Remedial Alternative Areas and Volumes ...................................................................................... 91

Table 25. Summary of Remedial Alternative Costs ($Millions) .................................................................... 92


ix

Table 26. Applicable or Relevant and Appropriate Requirements, LDW Superfund Site ........................... 107

Table 27. Selected Remedy RAO 3 RALs .................................................................................................... 121

Table 28. Remedial Action Levels, ENR Upper Limits, and Areas and Depths of Application ................... 125

Table 29. Cost Estimate Summary for Selected Remedy .............................................................................. 133

List of Figures

Figure 1. Lower Duwamish Waterway and Early Action Areas ....................................................................... 3

Figure 2. Potential Scour Areas and Estimated Net Sedimentation Rates ...................................................... 18

Figure 3. Conceptual Site Model. ................................................................................................................... 20

Figure 4. PCB Distribution in Surface Sediment ............................................................................................ 24

Figure 5. SMS Status in Surface Sediment ..................................................................................................... 27

Figure 6. LDW Areas with Parks and Habitat Restoration, Beach Play Activities, and Potential

Clamming. ..................................................................................................................................... 36

Figure 7. Conceptual Model for Baseline Human Health Risk Assessment ................................................... 37

Figure 8. Baseline Excess Cancer Risk and Noncancer Hazard Quotients for Consumption of Various

Seafood Species as a Function of the Number of Meals Consumed per Month ............................ 53

Figure 9. Baseline Noncancer Hazard Quotients and Excess Cancer Risk for the Seafood Consumption

RME Scenarios .............................................................................................................................. 54

Figure 10. Baseline Excess Cancer Risk for the Direct Sediment Contact RME Scenarios ........................... 55

Figure 11. Conceptual Models for the Ecological Risk Assessment .............................................................. 60

Figure 12. Recovery Category Areas .............................................................................................................. 85

Figure 13. Areas Addressed by LDW Cleanup Alternatives in the FS ........................................................... 86

Figure 14. Time to Achieve Risk Reduction for All Alternatives .................................................................. 93

Figure 15. Comparison of Total PCB Excess Cancer Risks and Noncancer HQs for Seafood Consumption

Calculated using LDW Baseline, Model-predicted, and Target Tissue Concentrations ............. 103

Figure 16. Excess Cancer Risks and Noncancer HQs for Seafood Consumption Calculated Using Target

Tissue Concentrations .................................................................................................................. 104

Figure 17. Recovery Category 1 and Potential Tug Scour Areas in LDW .................................................... 136

Figure 18. Selected Remedy .......................................................................................................................... 137

Figure 19. Intertidal Areas – Remedial Technology Applications................................................................. 139

Figure 20. Subtidal Areas – Remedial Technology Application ................................................................... 141

Figure 21. Intertidal and Subtidal Areas – Natural Recovery Application .................................................... 143

Figure 22. Intertidal Areas - Remedial Action Levels Application ............................................................... 144

Figure 23. Subtidal Areas – Remedial Action Levels Application ................................................................ 145


x

[This page intentionally left blank for double-sided printing.]


xi

Acronyms and Abbreviations

µg/kg micrograms per kilogram

ARAR applicable or relevant and appropriate requirement

ASTDR Agency for Toxic Substance and Disease Registry

AWQC Ambient Water Quality Criteria

BA biological assessment

BCM bed composition model

BiOp biological opinion

BEHP bis(2-ethylhexyl)phthalate

C combined technology

CAD contained aquatic disposal

CDI chronic daily intake

CERCLA Comprehensive Environmental Response, Compensation, and Liability Act

CFR Code of Federal Regulations

COC contaminant of concern

cPAH carcinogenic polycyclic aromatic hydrocarbon

COPC contaminants of potential concern

CSL cleanup screening level (SMS)

CSO combined sewer overflow

CT central tendency

CWA Clean Water Act

cy cubic yard

DMMP Dredged Material Management Program

DRCC/TAG Duwamish River Cleanup Coalition/Technical Advisory Group

dw dry weight

EAA Early Action Area

Ecology Washington Department of Ecology

EFDC Environmental Fluid Dynamics Code

EJ environmental justice

ELCR excess lifetime cancer risk

ENR enhanced natural recovery

EPA U.S. Environmental Protection Agency

EPC exposure point concentration

ERA ecological risk assessment

ESA Endangered Species Act

ESD explanation of significant differences

FS feasibility study

FWM food web model

HEAST Health Effects Assessment Summary Tables

HH human health

HHRA human health risk assessment


xii

HI hazard index

HPAH high molecular weight polycyclic aromatic hydrocarbon

HQ hazard quotient

HTLS higher trophic level species

HWTR Hazardous Waste Toxicity Reduction

I informational

IC institutional control

IRIS Integrated Risk Information System

LAET Lowest Apparent Effects Threshold

LDW Lower Duwamish Waterway

LDWG Lower Duwamish Waterway Group

LOAEL lowest observed adverse effects level

LPAH low molecular weight polycyclic aromatic hydrocarbon

mg/kg milligrams/kilogram (parts per million)

MLLW mean lower low water

MHHW mean higher high water

MNR monitored natural recovery

MOA memorandum of agreement

MOU memorandum of understanding

MTCA Model Toxics Control Act

MVUE minimum variance unbiased estimate

na, n/a, NA not applicable

NCP National Oil and Hazardous Substances Pollution Contingency Plan

nc not calculable or cannot be calculated

nd not detected

NEJAC National Environmental Justice Advisory Council

NEPA National Environmental Policy Act

ng/kg nanograms per kilogram (parts per trillion)

ng/L nanograms per liter

NOAA National Oceanic and Atmospheric Administration

NOAEL no observed adverse effects level

NPDES National Pollutant Discharge Elimination System

NTR National Toxics Rule

OC organic carbon

OM&M operation, maintenance, and monitoring

PCB polychlorinated biphenyl

PQL practical quantitation limit

PRG preliminary remediation goal

PRP potentially responsible party

PTW principal threat waste

R removal emphasis

RAL remedial action level


xiii

RAO remedial action objective

RBTC risk-based threshold concentration

RCRA Resource Conservation and Recovery Act

RCW Revised Code of Washington

RfD reference dose

RI remedial investigation

RL reporting limit

RME reasonable maximum exposure

RM river mile

ROC receptor of concern

ROD Record of Decision

SCO sediment cleanup objective (SMS)

SCWG Source Control Work Group

SD storm drain

SF slope factor

SMS Sediment Management Standards

SQS Sediment Quality Standard

STM sediment transport model

SVOC semivolatile organic compound

SWAC spatially-weighted average concentration

TBT tributyltin

TEQ toxic equivalent

TI technical impracticability

TMDL total maximum daily load

TOC total organic carbon

TRV toxicity reference value

TSCA Toxic Substances Control Act

TSS total suspended solids

UB upper bound

UCL95 upper confidence limit on the mean with 95% confidence

UECA Uniform Environmental Covenants Act

UL upper limit

USACE U.S. Army Corps of Engineers

U.S.C. United States Code

VOC volatile organic compound

WAC Washington Administrative Code

WDOH Washington Department of Health

WPCA Water Pollution Control Act

WQS Water Quality Standards

ww wet weight


xiv

[This page intentionally left blank for double-sided printing.].


1

Part 2 Decision Summary The Decision Summary provides an overview of the contamination present in the Lower Duwamish

Waterway (LDW) and the associated risks to human health and the environment, the cleanup alternatives

considered, and the U.S. Environmental Protection Agency’s (EPA’s) Selected Remedy to address these

risks. It also explains how the Selected Remedy fulfills statutory and regulatory requirements.

1 Site Name, Location, and Brief Description The LDW Site, located south of downtown Seattle, Washington, extends over the northern 5 miles of the

Duwamish River to the southern tip of Harbor Island (Figure 1), and includes upland sources of

contamination as well as the waterway. The southernmost portion of the site is located in Tukwila,

Washington. The Site was listed on the National Priorities List on September 13, 2001. The EPA

identification number for the Site is WA00002329803. Although the Site is not divided into operable

units, EPA and Washington State Department of Ecology (Ecology) have divided lead-agency

responsibility for addressing it. EPA has the lead for the In-waterway Portion to which this Record of

Decision (ROD) is addressed, and Ecology has the lead for upland source control.

The LDW and adjacent upland areas have served as Seattle’s major industrial corridor since the LDW

was created by widening and straightening much of the Duwamish River in the early 1900s. The

Duwamish River flows north through Tukwila and Seattle, splitting at the southern end of Harbor Island

to form the East and West Waterways, which discharge into Elliott Bay in Seattle, Washington. The In-

waterway Portion of the LDW Site addressed in this ROD extends for approximately 5 miles from the

area around the Norfolk Combined Sewer Overflow/Storm Drain (CSO/SD)2 at the southern end of the

Site at river mile (RM) 5 to the southern tip of Harbor Island at RM 0 (Figure 1). In total, the LDW

includes approximately 441 acres of intertidal and subtidal habitats. The average width of the LDW is

440 feet. Because this ROD does not address control of upland sources, and because the remedial

investigation/feasibility study (RI/FS) did not investigate the nature or extent of upland contamination,

this ROD does not define or bound the upland portion of the Site. To control contamination sources to

the In-waterway Portion of the Site, Ecology, in coordination with EPA, has identified the immediate

Source Area (see Section 2.4) which encompasses a total area of approximately 32 square miles, and

within which Source Control Strategy activities will take place (see Section 4.2).

The overall strategy for addressing contamination and the associated risks in the LDW and surrounding

watershed includes three components: 1) early identification and cleanup of the most contaminated areas

in the waterway, referred to as Early Action Areas (EAAs); 2) controlling sources of contamination to the

waterway; and 3) cleanup of the remaining contamination in the waterway, including long-term

monitoring to assess the success of the remedy in achieving cleanup goals. This ROD presents EPA's

Selected Remedy for component 3, cleanup of the In-waterway Portion of the Site. EPA is the lead

agency and Ecology is the support agency for component 3. Progress on component 1, cleanup of the

EAAs, is described in Section 2. For component 2, source control, Ecology’s proposed source control

activities are described in their draft Lower Duwamish Waterway Source Control Strategy (Source

Control Strategy) (Ecology 2012). EPA is the support agency for component 2.

2 The Norfolk CSO/ SD also serves as an emergency overflow for City pump station number 17. For brevity, it is

called CSO/SD in this document.


2

The Selected Remedy is intended to be the final remedy for the In-waterway Portion of the Site to be

implemented after cleanup in the EAAs has been completed, source control sufficient to minimize

recontamination (see Section 4.2) has been implemented, additional sampling and analysis has been

conducted, and design of the remedy has been completed.

2 Site History and Enforcement Activities

This section provides background information on past activities that have led to the current contamination

at the Site, and federal and state investigations and cleanup actions conducted to date under the

Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) and other

authorities.

2.1 Site History and Sources of Contamination

Most of the upland areas adjacent to the LDW have been industrialized since the early 1900s. This section

provides a brief history of industrial activities in the Duwamish valley and sources of contamination to the

Duwamish.

1900 – 1935: The start of industrial and commercial activity along the Duwamish coincided with the

dredging and straightening of the waterway in the early 1900s. Prior land use in the Duwamish valley

was primarily agricultural. Early activities included operation of sawmills, lumber yards, wood treatment

facilities, cement and brick companies, steel mills and foundries, and marine construction. Early facilities

include the Georgetown Steam Plant (built in 1906), The Boeing Company (Boeing) Plant 1 airplane

manufacturing (1917), and King County Airport/Boeing Field (1928).

1935 – 1955: Industrial use of the Duwamish accelerated with the onset of World War II. Many of the

existing industries (e.g., airplane and steel manufacturing) grew to support the war effort, and new

enterprises such as drum recycling and chemical production appeared. Boeing Plant 2 was constructed

from 1937 to 1940. Waste disposal practices in the 1950s and earlier included local landfills for solid

waste, soil infiltration for liquid waste, and direct disposal of liquid and solid waste into the waterway. A

primary treatment sanitary sewage facility on Diagonal Way was opened in 1938.

1955 – present: By 1955, sawmills, lumberyards and brick manufacturing facilities had virtually

disappeared. Current industrial uses include shipyard operations; manufacturing (airplane, cement, and

chemical, e.g., paint, glue, resin, and wood preservatives); cargo storage and transport; metal

manufacturing and recycling; and petroleum storage. A sanitary sewer system was constructed in the

1970s to serve both sides of the waterway. The landfills near the Duwamish are now inactive.

Historically, hazardous substances from upland industrial activities entered the environment through

spills, leaks, dumping, and other inappropriate management practices. Contamination entered the LDW

through a variety of pathways, including discharge through pipes, surface water or groundwater; dumping

materials directly into the waterway; or soil erosion.

Although waste disposal practices have improved considerably, legacy contamination continues to

threaten human health and the environment. At the same time, ongoing sources of contaminants are

present in this urban watershed, and pathways for both old and new contaminant sources continue to

transport contaminants to and from the LDW.


3

Figure 1. Lower Duwamish Waterway and Early Action Areas


4

Currently, contaminants enter the LDW through: direct discharges (stormwater, combined sewer

overflows, industrial wastewater); surface runoff sheet flow; groundwater discharges; erosion/leaching of

contaminated soils; in-water spills, dumping, leaks and inappropriate management practices; waterway

operations and traffic; atmospheric deposition; and transport of contaminated sediments (within the LDW

and from the upstream Green/Duwamish River watershed). Ecology’s draft Source Control Strategy (2012)

provides detailed information about pollution sources and pathways to the LDW.

Ecology’s investigations indicate that addressing the direct discharge and stormwater pathways for

contaminants is a higher priority than addressing the groundwater pathway. However, the groundwater

pathway is being evaluated and addressed as it is encountered during source control activities. For example,

EPA and Ecology have currently identified four groundwater plumes of primarily volatile organic

compounds (VOCs). Ecology has identified four facilities with polychlorinated biphenyls (PCBs) in the

groundwater. Three additional facilities have significant arsenic concentrations attributed to past industrial

activities that occurred on those facilities. Several of these and other facilities have groundwater

contaminated with carcinogenic polycyclic aromatic hydrocarbons (cPAHs) and metals. EPA and Ecology

are currently investigating and remediating all of these facilities. Ecology will continue to investigate and

address sources of contamination to the LDW, including groundwater, as part of its source control efforts.

2.2 Previous Investigations

Numerous investigations have been conducted to determine the nature and extent of contamination in the

LDW. Early studies included waterway-wide investigations of the LDW by the National Oceanic and

Atmospheric Administration (NOAA) and EPA in 1997 and 1998, respectively. At least 25 smaller,

location-specific investigations have been conducted by King County, the City of Seattle, Boeing, and other

private entities.

In December 2000, the City of Seattle, King County, the Port of Seattle, and Boeing, collectively known as

the Lower Duwamish Waterway Group (LDWG), voluntarily entered into an Administrative Order on

Consent with EPA and Ecology that requiried LDWG to conduct a remedial investigation/feasibility study

(RI/FS) pursuant to both CERCLA and the Model Toxics Control Act (MTCA) to investigate the nature

and extent of contamination and develop remedial alternatives for the In-waterway Portion of the Site.

During the RI, LDWG analyzed available data from numerous investigations conducted prior to 2000,

collected extensive additional data, conducted human health and ecological risk assessments, and identified

areas of greater contamination to be considered for early cleanup. The Final Lower Duwamish Waterway

Remedial Investigation Report (RI Report; LDWG 2010) was completed in 2010. The RI included an

assessment of risks to human health and the environment posed by the contamination. In the Final Lower

Duwamish Waterway Feasibility Study (FS Report; LDWG 2012a), completed in 2012, LDWG developed

alternatives for cleanup of the In-waterway Portion of the Site. After completion of the FS, LDWG

developed two supplemental memoranda (LDWG 2012b, 2013) which consider refinements to the

alternative under consideration for EPA’s preferred alternative in its 2013 Proposed Plan. EPA and

Ecology jointly provided oversight for the RI/FS.


5

2.3 Cleanup Activities Planned and Completed to Date

King County completed two cleanups of contaminated sediment in the LDW before and a few years after

the start of the RI/FS to partially implement the requirements of a 1991 CERCLA Natural Resource

Damages Consent Decree to address contamination from CSOs in Elliott Bay and the LDW:

In 1999, King County dredged 5,190 cubic yards (cy) of sediments contaminated with PCBs

outside the Norfolk CSO/SD. The area was then backfilled. (A small area of PCB-contaminated

sediments inshore of this cleanup was excavated [60 cy] and capped by Boeing under Ecology’s

Voluntary Cleanup Program in 2003.)

In 2003 and 2004, King County dredged (68,000 cy) and capped a 7-acre area around the

Duwamish/Diagonal CSO/SD. The contaminants of concern (COCs) that triggered this action were

PCBs, mercury, bis(2-ethylhexyl)phthalate (BEHP), and butyl benzyl phthalate. In 2005, a 6-inch

layer of clean sand was placed over an additional area where PCBs remained elevated after the

2003-2004 cleanup was completed.

The first phase of the RI identified areas with high levels of contamination for consideration for early

cleanup. Five EAAs were selected for action by EPA and Ecology, including the two King County

cleanups described above. Three of the cleanups have been completed (the King County cleanups described

above, and the Slip 4 cleanup described below), and two more will be completed (also described below)

before the Selected Remedy described in this ROD is implemented. Together, the cleanups at these five

EAAs (Figure 1) cover 29 acres, and address some of the highest levels of contamination found in the

LDW. Completion of the EAA cleanups will reduce the LDW-wide spatially area-weighted average surface

sediment PCB concentration by an estimated 50%.

EPA is conducting or has completed cleanups in the following areas.

Completed cleanups:

Slip 4: Approximately 10,000 cy of PCB-contaminated sediments were dredged and 3.4 acres were

capped with clean sand, gravel, and granular activated carbon amended filter material, from

October 2011 through January 2012, by the City of Seattle (with participation by King County)

under an Administrative Settlement Agreement and Order on Consent (consent order) for a

CERCLA removal. The first Long-Term Monitoring Data Report (for Year 1, 2013) was approved

in January 2014.

Ongoing cleanups:

Terminal 117: Soils on the upland portion of T-117 industrial property (referred to as “Upland”)

with elevated concentrations of PCBs and other contaminants were removed by the Port of Seattle

with EPA oversight pursuant to separate 1999 and 2006 CERCLA consent orders. Cleanup of eight

residential yards and an alleyway within the Study Area was completed in 2013, and cleanup of

streets and rights of way, and installation of a permanent stormwater system, should be completed

by the City of Seattle in 2015. Cleanup of the Upland soils and off-shore contaminated sediments

was mostly completed in 2013-2014 by the Port of Seattle, with a small portion of the bank work

remaining and scheduled for completion in winter 2014. All of the removal work conducted in the

Study Area is being done by the Port and City under a June 2011 CERCLA consent order.

Boeing Plant 2/Jorgensen Forge: Following completion of sufficient source control actions within

the upland portion of the Jorgensen Forge facility and concurrent implemenation of interim


6

corrective measures and development of a Corrective Measures Study for the upland Boeing Plant

2 facility, cleanup in areas of sediment contamination in the banks and offshore of the adjacent

Boeing Plant 2 and Jorgensen Forge facilities started in 2013. Although these areas are identified as

one EAA, they are being addressed as separate actions pursuant to separate EPA decision

documents and consent orders under different laws that require implementation coordination.

Boeing Plant 2 contaminated sediments are being addressed by Boeing under a 1994 Resource

Conservation and Recovery Act (RCRA) Consent Order. Sediments contaminated with metals and

other hazardous substances at Jorgensen Forge are being cleaned up under a 2012 CERCLA

removal consent order by Earle M. Jorgensen, a former owner of the facility. EPA anticipates

completion of both of these early actions in 2015.

The following timeline provides a summary of LDW activities to date.

Lower Duwamish Waterway Timeline

1999 Sediment cleanup was completed at the Norfolk CSO/SD

2000 A CERCLA and MTCA Consent Order was issued by EPA and Ecology requiring LDWG to conduct the RI/FS

2001 LDW was listed as a Superfund site on the National Priorities List

2002 LDW was listed by Ecology as a cleanup site under MTCA

EPA and Ecology signed a Memorandum of Understanding (MOU) designating EPA as the lead for in-waterway

cleanup, and Ecology as the lead for source control. The MOU was revised in 2004, and was revised again in

2014 when the title was changed to Memorandum of Agreement (MOA).

Ecology initiated the Source Control Work Group

2003 The LDW Phase 1 RI was completed, and additional cleanup was conducted at the Norfolk CSO/SD

2004 Ecology issued its Source Control Strategy

2005 Sediment cleanup was completed at the Duwamish/Diagonal CSO/SD

2006 EPA issued an Action Memorandum (cleanup plan) for the Slip 4 EAA

2010 The Final LDW RI was completed

2010 EPA issued an Action Memorandum for the Terminal 117 EAA

2011 EPA issued a RCRA corrective action Final Decision (cleanup plan) for Boeing Plant 2 sediment

EPA issued an Action Memorandum for Jorgensen Forge sediments and shoreline bank soils

2012 Sediment cleanup was completed at the Slip 4 EAA

The Final LDW FS was completed

2013 Cleanup started at Boeing Plant 2 and Terminal 117

2014 Cleanup started at Jorgensen Forge


7

2.4 Source Control Investigations and Actions Completed to Date

Ecology is the lead agency for identifying direct and indirect sources of contaminants to the In-waterway

Portion of the Site. Ecology uses its regulatory authority and works with other governments that have

regulatory authority (EPA, King County, and City of Seattle), also referred to as the Source Control Work

Group (SCWG), to control ongoing sources to the extent possible. The SCWG began its work in 2002, with

the goal of identifying, prioritizing, and controlling sources of contamination to the LDW to support

remedy selection in this ROD.

Members of the SCWG performed numerous investigations to identify ongoing sources of contaminants.

These include:

Compiling a waterway-wide summary of potential sources and investigating those potential

sources.

Developing Source Control Action Plans for each of the Source Control Action Areas that drain to

the LDW. Ecology (with the SCWG) has identified 24 distinct Source Control Areas, and has

completed Source Control Action Plans for all of them. Each plan identifies the authorities, tools,

and milestone accomplishments for controlling the sources and identifies criteria or other goals that

determine effectiveness and completeness of source control actions within each drainage basin.

Tracing sources by sampling for contaminants in solids within storm drains and catch basins. This

helps identify facilities where historical, unidentified, or illegal disposal of contaminants has

occurred or is occurring, making the facility an active source of contaminants affecting the LDW.

Investigating and addressing contamination at upland facilities, including those contributing

contamination to the LDW via groundwater, stormwater, soil erosion, or air deposition.

Developing and implementing other studies to identify ongoing sources, including: inputs to the

Green/Duwamish River; inputs due to outfalls and other lateral sources; and inputs of PCBs in or

from building materials in the source area.

Additionally, Ecology administers the Clean Water Act’s National Pollutant Discharge Elimination System

(NPDES) permitting program in Washington State. Ecology, with the other members of the SCWG, has

made substantial progress in finding, investigating, and controlling both historical and ongoing sources to

the LDW, though more work remains. The summary on the next page highlights numerous ongoing LDW

source control actions. More detailed information about the source control studies and work to date can be

found on Ecology’s website at: http://www.ecy.wa.gov/programs/tcp/sites_brochure/lower_duwamish/

lower_duwamish_hp.html.

2.5 Enforcement Activities

In addition to the CERCLA and MTCA RI/FS and removal orders and RCRA corrective action orders

discussed above, EPA conducted activities to identify additional potentially responsible parties (PRPs) who

may have contributed to contamination in the LDW. Information request letters were sent to 276 parties

from 2006 to 2013. General notice letters were sent to 113 parties in November 2012. The general notice

letters provided notification of the recipients’ opportunity to comment on the Proposed Plan.

http://www.ecy.wa.gov/programs/tcp/sites_brochure/lower_duwamish/lower_duwamish_hp.html



8

Summary of Source Control Actions To Date

All the source control work conducted to date and summarized below involved one or more of the following: source control investigations, site

assessment and cleanup, inspections, source tracing, sampling, and monitoring. One hundred ninety-six confirmed or suspected

contaminated upland facilities have been identified within the LDW drainage basin, although only some of those are sources of contaminants

to the LDW.

Thirteen facilities along or near the LDW are under agreed orders for investigation and cleanup administered by Ecology’s Toxic Cleanup

Program:

– Jorgensen Forge Corporation – North Boeing Field/Georgetown Steam Plant

– 8801 East Marginal Way (former Paccar site) – Fox Avenue/Great Western Chemical

– South Park Landfill – Glacier NW/Reichhold

– Crowley Marine Services – Duwamish Shipyard

– Industrial Containers/Trotsky/Northwest Cooperage – Douglas Management Properties

– Boeing Isaacson-Thompson – Port of Seattle Terminal 115 North

– Duwamish Marine Center

Five additional facilities in the LDW source area are under agreed orders for investigation and cleanup administered by Ecology’s Hazardous

Waste Treatment and Reduction (HWTR) program:

– Art Brass Plating – Blaser Die Casting

– Capital Industries – General Electric — Dawson Street Plant

– Philip Services Georgetown

Ecology has conducted site investigations at:

– South Park Marina (former A and B Barrel) – Basin Oil

– Washington State Liquor Control Board Warehouse – Douglas Management Company

– Industrial Container Services (formerly Northwest Cooperage)

Four voluntary cleanups under MTCA are occurring or have been completed at:

– Boeing Developmental Center – Port of Seattle Terminal 106/108

– General Services Administration — Federal Center South – City of Seattle 7th Ave Pump Station

(Approximately ten other voluntary cleanups are ongoing or completed within the LDW Source Area at facilities not adjacent to the LDW)

Eight facilities along or near the LDW are under an EPA cleanup process:

– Boeing Plant 2 (RCRA) – Jorgensen Forge shoreline (CERCLA)

– Rhône-Poulenc (RCRA) – Port of Seattle Terminal 117 (CERCLA)

– Boeing Electronics Manufacturing Facility (CERCLA) – Tully’s/Rainier Commons (Toxic Substances Control Act)

– 24” stormwater outfall Boeing/Jorgensen property line (CERCLA)

– North Boeing Field/King County International Airport Storm Drain Treatment System (CERCLA)

In addition:

Between 2003 and December 2013, the City of Seattle and King County have completed more than 3,400 inspections at nearly

1,500 businesses in the LDW area. In addition, they have collected more than 1,025 sediment samples from storm drains and

combined sewer systems to help identify and characterize sources discharging to the municipal storm and wastewater collection

systems.

In 2008, Ecology signed an interagency agreement with the City of Seattle to expand source tracing sampling. As part of this

agreement, Seattle Public Utilities installed twenty additional sediment traps in the LDW study area, including areas on King County

International Airport and unincorporated King County.

From October 2009 through December 2013, Ecology’s Lower Duwamish Urban Waters Initiative inspection team completed 260

water quality inspections and 321 hazardous waste inspections.

Approximately 104 facilities in the LDW drainage basin have NPDES permits from Ecology; approximately 94 facilities are regulated

under a general industrial stormwater permit; two active facilities have individual industrial wastewater discharge permits; two

facilities operate under a general permit for boatyards; and six facilities operate under a general permit for sand and gravel facilities.

Four local governments have municipal separate stormwater general discharge permits (Phase I for the City of Seattle and King

County, and the Port of Seattle as a secondary permittee; and Phase II Western Washington for the City of Tukwila).

Two local governments (the City of Seattle and King County) have individual discharge permits for their combined sanitary sewer

and stormwater systems. Both entities have prioritized CSO control projects to address discharges to the LDW.

For comprehensive accounts, and up to date information, check the most recent Source Control Status Reports on Ecology’s website at

http://www.ecy.wa.gov/programs/tcp/sites_brochure/lower_duwamish/lower_duwamish_hp.html.



9

3 Community and Tribal Participation

This section summarizes the community involvement and Tribal consultation activities performed by

EPA and Ecology during the RI/FS and the remedy selection process. In 2002, EPA and Ecology

developed a community involvement plan to promote meaningful involvement of the public during the

investigation and cleanup of the LDW. This plan was developed based on interviews with community

members and identified stakeholders. Throughout the RI/FS, EPA and Ecology have:

regularly held public meetings and have attended community and advisory group meetings;

held quarterly stakeholder meetings to provide updates on the RI/FS, cleanup of the EAAs, and

source control activities;

consistently sought input from the Tribes, community groups, and natural resource agencies

when reviewing and commenting on sampling plans, the human health and ecological risk

assessments, and other RI/FS documents;

sent fact sheets to inform the community about Site progress;

provided opportunities for public comment on the RI and FS Reports;

provided information about EPA's work at the Site at annual community festivals; and

provided updates at neighborhood meetings.

EPA and Ecology used input from a 2010 public review of the draft FS to finalize the FS and develop the

Proposed Plan. EPA provides technical assistance grants to the community advisory group for the Site,

the Duwamish River Cleanup Coalition/Technical Advisory Group (DRCC/TAG). This organization

reviews information about the Site and shares it with community members. EPA plans to continue to

coordinate with Ecology and engage with the community throughout design, construction, and long-term

monitoring of the remedy, including any potential modifications to the remedy.

The LDW is one of the locations of the Muckleshoot Tribe’s commercial, ceremonial, and subsistence

fishery for salmon, as part of its usual and accustomed fishing area. The Suquamish Tribe actively

manages aquatic resources north of the Spokane Street Bridge, just north of the LDW study area.

Consideration of how Tribal members may be exposed to contaminants in the LDW while engaging in

seafood harvest activities has been a primary factor shaping the assessment of human health risks. The

Tribes, as sovereign nations, have engaged in government to government consultations with EPA on the

cleanup process and decisions. The Tribes have also broadly and actively participated in meetings

determining the course of the cleanup to date. EPA plans to continue to consult with the Muckleshoot and

Suquamish Tribes throughout design, construction, and long-term monitoring of the remedy, including

any potential modifications to the remedy.

In conjunction with the FS, in response to comments on the 2010 draft FS, an environmental justice

analysis (EJ Analysis) for the LDW was conducted and a draft was appended to the Proposed Plan (EPA

2013a) as Appendix B and made available for public comment concurrently with the Proposed Plan. EPA

defines environmental justice as “the fair treatment and meaningful involvement of all people regardless

of race, color, national origin, or income with respect to the development, implementation, and


10

enforcement of environmental laws, regulations, and policies.” The purposes of the EJ Analysis were

1) to screen for EJ concerns, and 2) to identify disproportionate adverse impacts from the cleanup

alternatives and the Preferred Alternative in the Proposed Plan and, if found, provide recommendations to

mitigate such impacts. The information and recommendations from the EJ Analysis have been considered

in the development of this ROD.


11

4 Scope and Role of the Response Action

The Selected Remedy is the third and final part of an overall strategy for addressing contamination in the

LDW Site that includes: 1) early identification and cleanup of EAAs to address the most contaminated

areas in the waterway; 2) controlling sources of contamination to the waterway; and 3) cleanup of the

remaining contamination in the waterway, including long-term monitoring to assess the success of the

remedy in achieving cleanup goals. These three components together are designed to address the areal

extent of contamination at the Site, including sediment contamination, resident seafood tissue (edible

portions of fish and shellfish) concentrations, and water quality within the waterway, to the extent

practicable. As described below, the Selected Remedy is intended to be the final remedy for the In-

waterway Portion of the Site to be implemented after completion of additional sampling during the design

phase of remedy implementation (remedial design), and after implementation of the EAA cleanups and

sufficient source control to minimize recontamination.

4.1 Component 1: Early Identification and Cleanup of EAAs

The first phase of the LDW RI included identification of the most contaminated areas of the waterway for

consideration as EAAs. Section 2.3 describes progress to date on cleaning up the selected EAAs. Cleanup

alternatives, costs, and outcomes identified in this ROD assume completion of the EAA cleanups, all of

which are scheduled for completion by the end of 2015.

EPA has reviewed the EAA cleanup actions being performed under EPA Consent Orders and has

determined that the completed Slip 4 EAA is consistent with the Selected Remedy and requires no further

active remediation. The other selected EAAs are similarly expected to require no further active

remediation if they achieve their stated objectives. Nevertheless, as with the rest of the LDW, all the

EAAs will be subject to performance review to assure that human health and the environment are being

protected. In conducting performance reviews, EPA will review the Institutional Controls Plans and long-

term monitoring plans for all EAAs and will require that the EAAs be incorporated into plans for the rest

of the LDW as necessary to make them consistent with the Selected Remedy in the ROD. For the

cleanups conducted under the 1991 Natural Resource Damages Consent Decree (Duwamish/Diagonal

CSO/SD and Norfolk CSO), EPA will conduct a review during the remedial design phase to determine

whether any additional work is needed to make these cleanup actions consistent with the Selected

Remedy in this ROD.

4.2 Component 2: Controlling Sources of Contamination

As a general principle, EPA seeks to control sources of contamination early when managing risks at

hazardous waste sites. Sources of contaminants in LDW surface water and sediments include stormwater

carrying the contaminants of concern via CSOs, stormwater drains, and other point and non-point source

discharges; upland facilities or source areas with contaminants discharging to the LDW via groundwater,

surface water, or erosion of contaminated soils; and atmospheric deposition of COCs. Ecology and the

SCWG have performed extensive investigations and initiated multiple actions to address known sources

of contaminants. Section 2.4 and Ecology’s Source Control Strategy (Ecology 2012) provide more

information on how Ecology as the lead agency for source control is leading this important component of

the overall LDW remediation.


12

An objective of source control is to find and sufficiently control sources before conducting in-waterway

remediation and thereby prevent or minimize recontamination after the cleanup is completed. EPA and

Ecology will coordinate to find and sufficiently control sources of sediment contamination. Based on

communications with Ecology, EPA anticipates that the process for determining whether source control is

sufficient to begin in-water work will be fully described in Ecology's final Source Control Strategy, to be

released in 2015. The focus of this work is to control sources sufficiently such that recontamination above

the benthic SCO criteria and human health remedial action levels (RALs) (see Section 13.2 for an

explanation of SCOs and RALs) is unlikely. Ecology plans to divide the waterway into three sections for

the purposes of prioritizing source control activities, sequencing activities from the upstream to

downstream sections of the LDW. Baseline and/or remedial design data will be necessary prior to

conducting a sufficiency evaluation. After an evaluation, Ecology will provide EPA a recommendation

whether or not to proceed with in-water remedial actions based on the status of source control in the

immediate vicinity of the planned action. Upon EPA’s concurrence that source control is sufficient, active

in-waterway sediment remediation can begin. This will prevent or minimize the likelihood that sediments

will be recontaminated at levels that trigger additional active in-waterway sediment remediation (see

Section 13). The coordination of the source control and in-waterway cleanup activities has been

established in a Memorandum of Agreement (MOA). The MOA provides a broad framework for

organizing the work of the federal, state, and local agencies under various legal authorities and describes

how EPA and Ecology will coordinate sequencing of source control and sediment remedial actions. As

stated earlier, Ecology’s draft Source Control Strategy (Ecology 2012) provides a broad framework for

organizing the work of the federal, state, and local agencies under various legal authorities.

Overall, source control activities are expected to occur on two scales: 1) the immediate source area to the

LDW where source control activities are focused on controlling sources and pathways of contamination to

LDW sediments to prevent or minimize the likelihood that sediments will be recontaminated at levels that

trigger additional active in-waterway sediment remediation; and 2) the larger watershed where source

control activities are focused on regional efforts to address toxics that are present at ubiquitous

concentrations in sediments, surface water, and stormwater, and via air deposition.

The final Source Control Strategy is expected to reference a Pollutant Loading Assessment (PLA; EPA

and Ecology 2014) for the watershed to address the activities in the larger watershed (described in item 2

above). The PLA is needed to understand the relationship of water, sediment, and fish tissue quality to the

overall health of the Green/Duwamish watershed. The goal of this assessment is to determine ways to

reduce ongoing sources of pollution in the watershed. The PLA will help evaluate the relative importance

of various sources of pollution to the watershed and inform targets and strategies for reducing those

sources of pollution. The PLA tool will assess relative contribution of pollutants from various sources and

pathways and will provide information to help prioritize source control activities in the watershed. The

PLA will support and enhance Ecology and EPA’s current efforts to clean up the LDW. The LDW in-

waterway cleanup is expected to significantly improve sediment quality in the LDW, but the success of

the cleanup relies in part on cleaner sediments from upstream depositing on the LDW over time. By

identifying strategies to reduce sources of pollution throughout the watershed, the PLA will help improve

the effectiveness of the in-waterway cleanup.

Ecology sought public comment on its 2012 draft final revision of its 2004 Source Control Strategy

concurrently with EPA’s public comment period for the Proposed Plan and is currently in the process of


13

revising the Source Control Strategy to address public comments. Ecology is currently working with the

federal, state and local source control agencies to develop Source Control Implementation Plans

(Implementation Plans) as part of the final Source Control Strategy. The Implementation Plans will

describe how each agency will conduct its various programs to address source control work for the LDW

source area. Ecology is currently requesting Implementation Plans from EPA, King County, and the City

of Seattle, and may request Implementation Plans from other entities in the future. EPA submitted its

draft Implementation Plan, which details its internal coordination process for source control projects

within the LDW, to Ecology in 2013. Ecology will develop its own Source Control Implementation Plan.

Ecology is expected to finalize its Source Control Strategy upon completion of the Implementation Plans,

which it anticipates completing in 2015. Once completed, these plans will be available on Ecology's web

site. In the unlikely event that timely and effective source control is not implemented, EPA may take

actions pursuant to CERCLA or other federal authority to ensure the implementation and protectiveness

of the Selected Remedy.

4.3 Component 3: In-Waterway Cleanup

The third element of the overall cleanup strategy for the Site, and the focus of this ROD, is the in-

waterway cleanup. The Selected Remedy addresses, to the extent practicable, contaminated sediments and

surface water below the mean higher high water (MHHW) level (in the LDW, MHHW is 11.3 feet above

the mean lower low water [MLLW] level)3 that are expected to remain after the EAA cleanup work

(component 1) is completed. Although the Selected Remedy does not directly address surface water, COC

concentrations in surface water will be reduced through implementation of source control and the

Selected Remedy. The active response actions selected in this ROD will be implemented after completion

of additional sampling and remedial design, and implementation of the EAA cleanups (component 1) and

sufficient source control (component 2) to minimize recontamination in any particular area within the

waterway.

The Selected Remedy described in this ROD is a final action that will be protective of public health and

the environment, as described in detail in Section 13.4. It is EPA’s expectation that COC concentrations

in sediment and in fish and shellfish tissue will have been reduced by 90% or more once the EAA

cleanups are complete, adequate source control has been implemented, the active cleanup portions of the

Selected Remedy have been implemented, and 10 years of monitored natural recovery have occurred after

completion of the active cleanup portions of the remedy.

Although one goal of the combined actions is to attain applicable or relevant and appropriate

requirements (ARARs), it may not be possible to attain all ARARs. Institutional controls (ICs) that limit

seafood consumption in quantitites that present unacceptable health risks will be needed for the

foreseeable future. The intent of the Selected Remedy is to reduce contaminant concentrations in

sediments, surface water, and fish and shellfish tissue to the extent practicable, and to minimize reliance

on fish and shellfish consumption advisories to reduce human exposure from ingestion of contaminated

resident fish and shellfish. If EPA determines, after implementation of the remedy and long-term

monitoring, that is it not technically practical to attain ARARs through the Selected Remedy or additional

3 The LDW has two high tides and two low tides each day. MLLW is the average lowest daily low-water height and

mean higher high water (MHHW) is the average highest daily high-water height, averaged over many years.


14

actions, EPA will consider waiving the ARAR in a future ROD Amendment or Explanation of Significant

Differences (ESD) (see Sections 8.2.1 and 13.4).

It is important to note that meeting the requirements of CERCLA supports, but is not the same as

attaining Washington's designated or existing uses under the Clean Water Act (CWA), as identified in

WAC 173-201A. This ROD addresses the In-waterway Portion of the Site. Ecology is implementing

other complementary actions, including actions under the CWA, to protect seafood consumers and

aquatic life within the LDW. The CWA addresses pollutants in the water column through various

mechanisms, including the NPDES permitting program under Section 402 of the CWA, which requires

that point sources not cause or contribute to water quality standards violations, and the water quality

standards and implementation plan program under Section 303 of the CWA. EPA delegated CWA

permitting authority in Washington State to Ecology, and Ecology issues NPDES permits in compliance

with CWA and state Water Pollution Control Act (WPCA) that authorize discharges to the LDW.

Washington State’s water quality assessment report prepared under section 303(d) and 305(b) of the

CWA identifies numerous impairments in the LDW, including failure to meet water quality standards for

approximately 40 different pollutants in sediment and fish tissue. The CWA requires that these

impairments be addressed through development of a total maximum daily load (TMDL) or through the

implementation of other pollution controls that will ensure that water quality standards are attained. EPA

acknowledges that in order to address wide-spread contamination in the water column, Ecology may need

to apply existing or revised water quality standards implementation tools. Changes to existing water

quality standards implementation tools would be subject to EPA review and approval under the CWA.

Over time, the integrated approach of CERCLA and longer-term clean water actions is expected to result

in attainment of applicable surface water quality criteria and uses designated under the CWA.


15

5 Site Characteristics This section summarizes information obtained through the RI/FS and other investigations conducted

before or during the RI/FS. It includes a description of the physical characteristics of the LDW, including

the results of the sediment transport model (STM) and the bed composition model (BCM), which are

described in Sections 5.1.2 and 5.1.3, respectively, and the overall conceptual site model (CSM),

described in Section 5.2. The CSM is supported in part by the food web model (FWM) used in the RI/FS

and described in Section 5.2.1. Section 5.3 provides information on the nature and extent of

contamination in the LDW, including concentrations of contaminants of concern (COCs) in the LDW as

well as background levels and levels that enter the LDW from upstream.

5.1 Physical Characteristics

The LDW, originally the natural meandering estuary at the confluence of the Green/Duwamish River

system and Elliot Bay, was modified in the early 1900s to become an engineered navigation channel for

commercial use, termed a waterway, from RM 0 to RM 4.7. The In-waterway Portion of the Site extends

from RM 0 to RM 5, encompassing approximately 441 acres. Most of the natural wetland habitat and

mudflat areas associated with the original Duwamish River estuary are no longer present as a result of the

waterway construction and subsequent upland development.

The central portion of the waterway is maintained as a federal navigation channel by the US Army Corps

of Engineers (USACE). The navigation channel is maintained at authorized navigable depths of 30 ft

below “mean lower low water” (-30 ft MLLW) from Harbor Island to the First Avenue South Bridge (RM

0 to 2), at -20 ft MLLW from the First Avenue South Bridge to Slip 4 (RM 2 to 2.8), and at -15 ft MLLW

from Slip 4 to the Upper Turning Basin (RM 2.8 to 4.7). Depths outside the navigation channel

immediately south of Harbor Island at the mouth of the waterway are as deep as -47 ft MLLW. To

maintain navigation depths, USACE dredges the upstream portion of the navigation channel every one to

three years. USACE typically dredges 2 ft below the authorized depths (“advanced dredging”) in order to

assure navigable depths are maintained until the next dredging event. The area typically dredged is the

Upper Turning Basin and downstream to approximately RM 4. In addition, private parties periodically

dredge berthing areas to maintain depths for their own purposes, typically shipping and marina uses.

Outside the navigation channel, the LDW banks are comprised of sloped subtidal embankments, shallow

subtidal and intertidal areas (including five slips along the eastern shoreline and three embayments along

the western shoreline), and Kellogg Island near the downstream end. The shoreline consists primarily of

hardened surfaces, including riprap, aprons for piers, and sheet-pile walls, with some beaches and

intertidal habitat remaining in isolated patches.

5.1.1 Surface Water Hydrology

Human activity has greatly influenced water and sediment movement in the LDW. Rivers that historically

flowed into the upstream Green River were diverted in the early 1900s, reducing the volume of water

entering the LDW by approximately 70%. Water flows are now managed approximately 65 miles

upstream by the Howard Hanson Dam, constructed in 1961. As a result, peak flows are much smaller with

maximum flows rarely exceeding 12,000 cubic feet per second (cfs). Average river flows are estimated to

be 1,340 cfs. In addition, the LDW has been widened and deepened to allow for navigation, resulting in

reduced velocities. The reductions in peak flows and velocities result in less erosion and more deposition

of sediments.


16

The LDW is a two-layer salt wedge estuary, with outflow that is mostly freshwater originating from the

Green/Duwamish River at the surface, and tidally-influenced salt water from Puget Sound entering the

LDW at the mouth of the waterway beneath it. The saltwater “wedge,” or interface between fresh water at

the surface and salt water at depth, is always present from RM 0 to RM 2.2, and is periodically present

between RM 2.2 and RM 4, depending on tide height and river flow. Between RM 4 and RM 5 freshwater

is usually predominant, although the saltwater wedge can extend through this area when the tide is high

and river flow is low. LDW tidal fluctuations average about 11 feet. The presence of the denser salt water

layer throughout much of the waterway forces the bulk of the freshwater to discharge upward toward the

surface of the waterway, thus reducing sediment erosion during high river flows.

5.1.2 Sediment Transport Model

A three-dimensional sediment transport model (STM; LDWG 2008) was developed to simulate water

flow and sediment erosion and deposition over a wide range of flow and tidal conditions to inform the

type of sediment cleanup technologies that would be appropriate for the area. The STM linked a sediment

transport model and a hydrodynamic model called Environmental Fluid Dynamics Code (EFDC) (see

Section 5.2.1). The STM estimated that, on average, more than 200,000 metric tons of sediment enters

the LDW each year. About 50% of the incoming sediment deposits within the LDW. The rest is exported

further downstream to Elliott Bay. Approximately 50% of the sediment that settles in the LDW is

removed by periodic navigational maintenance dredging. Thus, approximately 25% of the incoming

sediment remains in the LDW after dredging. The annual average amount dredged from the LDW by

USACE is 51,000 metric tons, mostly in the Upper Turning Basin.

Based on the STM, approximately 99% of the sediment entering the waterway is from upstream. The

other approximately 1% is directly discharged into the LDW via storm drains, CSO outfalls, and small

streams. Although direct discharges to the LDW only account for approximately 1% of the sediment load

to the LDW, the contaminant concentrations in these sediments are much higher than in the sediments

coming in from upstream. This often causes elevated contaminant concentrations in localized areas

around outfalls. Sources that may be contributing to these higher concentrations are being investigated

and addressed as part of source control (see Section 2.4).

Erosion and deposition4 rates predicted by the STM are illustrated in Figure 2

5. The STM results indicate

that, overall, there is more deposition than erosion of sediment in the LDW. The highest net

sedimentation rates in the LDW (up to 151.5 cm/year) occur in the area of the Upper Turning Basin from

RM 4 to RM 4.8. The Upper Turning Basin serves as a trap for much of the coarser fraction of the bed

load entering the LDW from upstream, and is dredged by USACE every few years. Throughout the

LDW, net sedimentation rates were generally greater than 1 cm/year in the subtidal areas and less than

1 cm/year in the intertidal areas. Due to channel morphology, some areas are more erosional and some are

more depositional. Erosion of the sediment bed by river flow (termed high-flow scour in Figure 2) is

limited, even during high-flow events. Most bed erosion due to high-flow scour is less than 10

centimeters (cm) in depth and the maximum estimated net erosion depth is 22 cm. Routine vessel

operations in shallow areas and berthing areas may cause localized propeller-wash scour (also known as

vessel scour or tug scour) to depths greater than 22 cm, but likely less than 60 cm; routine vessel

4 The combined or net effect of erosion and deposition is termed net sedimentation.

5 In addition, Figure 2 shows areas that have evidence of propeller-wash scour which could redistribute sediments

but were not included in the STM. See Section 9.2 and Table 22 for further discussion.


17

operations in the navigation channel are predicted to mix sediments to depths of 1-2 cm. Note that

routine operations do not include emergency operations or shoreline and shallow water maintenance

events, which occur sporadically in the waterway. The STM’s predictions are corroborated by sediment

contaminant concentration data collected in the same locations over time, which indicate that natural

recovery (as described in Section 9.3) is occurring in some areas of LDW.

5.1.3 Bed Composition Model

A different model, the bed composition model (BCM), was used to estimate future COC concentrations in

LDW sediments. The BCM used predictions of sediment movement from the STM, data on sediment

contaminant concentrations in the LDW and in sediment entering the LDW from the Green/Duwamish

River, and data on other sediment inputs to the LDW from ditches, streams, and municipal discharges in

the LDW basin. The BCM provided predictions of approximate future sediment contaminant

concentrations that would exist after implementation of each of the proposed cleanup alternatives.

As discussed in Section 10.1, the STM, the BCM, and empirical evidence were used in configuring and

evaluating the long-term effectiveness of remedial alternatives. They were used to evaluate whether the

sediment bed is stable (i.e., not subject to significant scour, erosion, and transport) and whether the net

sedimentation rate is sufficient for natural recovery (burial of contaminated sediments) to occur. If these

conditions are met in a given location, then monitored natural recovery (MNR) may be an applicable

response action to evaluate in one or more remedial alternatives. Conversely, if natural processes do not

appear to be effectively reducing contaminant concentrations in surface sediments, then only active

remedial measures have been considered.

Key inputs to the BCM were upstream, lateral (e.g., from stormwater and CSOs), and bed-load

contributions. Sensitivity testing of the BCM was performed as part of the FS to understand how model

inputs affect the results. Low-end, mid-range, and high-end values were established for each type of

input. The mid-range values were used as model inputs for developing and analyzing alternatives, and the

low-end and high-end values were used for analyzing model sensitivity. Also, smaller scale areas were

analyzed to evaluate local recovery potential and to assess whether empirical data and predictive models

agree. Several lines of evidence6 were combined to assess whether contaminated subsurface sediments are

stable and are unaffected by surface-sediment movements, and whether surface sediment contaminant

concentrations are expected to decrease over time.

6 Lines of evidence considered for these purposes included isotopic analysis in core samples, sediment transport

analysis, contaminant trend analyses, and evaluation of erosion potential.


18

Figure 2. Potential Scour Areas and Estimated Net Sedimentation Rates


19

5.1.4 STM and BCM Uncertainty and Sensitivity Analyses

As noted above, analyses of uncertainty in the STM and BCM and of their sensitivity were performed during

the RI and FS. The primary sources of uncertainty in the STM and BCM predictions are: 1) COC

concentrations in incoming sediments from upstream and lateral sources; 2) the rate of net sedimentation

(sediment burial) from incoming sediment loads (or net erosion); and 3) the potential for deep disturbances of

subsurface contaminated sediments by mechanisms not accounted for in the model, such as vessel scour

(propeller-wash or tug scour) and earthquakes.

While long-term projections of contaminant concentrations and the time to reach the lowest model-projected

concentrations must be viewed in light of the uncertain inputs, the STM, BCM, and food-web modeling, along

with a subsurface-disturbance analysis in the FS, provide sufficient basis for comparison of alternatives and

selection of the remedy. STM and BCM predictions for future waterway levels of contamination did not

directly incorporate disturbances to bed sediments from propeller-wash scour; therefore, imaging of the

waterway bed, interviews with waterway users, and sediment trend data were used to further refine

understanding of these areas. Areas identified as potential “propeller-wash scour areas” are shown in Figure 2;

however, they do not account for the possibility that an infrequent vessel scour event can occur virtually

anywhere in the waterway. An additional uncertainty analysis evaluated the potential increase in contaminant

concentrations due to infrequent scour events for each of the FS alternatives using PCBs as an indicator for all

other contaminants.

5.2 Contaminant Transfer Conceptual Site Model Figure 3 summarizes the pathways of contaminant transfer within the LDW. Multiple hazardous substances

have been and continue to be discharged into the LDW and remain in the water column and waterway

sediments. The sources of contamination and release/migration pathways are more fully described in

Ecology’s 2012 draft Source Control Strategy. Once contamination enters the waterway via the migration

pathways shown in Figure 3, the contaminants may be taken up by organisms, including fish and shellfish, and

bottom-dwelling organisms (also called benthic invertebrates). The consumption of these organisms by larger

fish, shellfish, and wildlife provides a mechanism for the contaminants to move from the sediment and water

through the food chain. This poses threats to human health and the environment when people and wildlife

consume resident fish and shellfish from the LDW. Fish such as salmon that only migrate through the LDW,

instead of being resident, do not take up enough contaminants while in the LDW to be a concern. People and

wildlife may also face risks from direct contact with contaminated LDW sediments. Section 7 describes

exposure pathways for humans and wildlife and the associated risks.


20

Figure 3. Conceptual Site Model.


21

5.2.1 Food Web Model

A food web model (FWM; LDWG 2010) based on a model initially developed by Arnot and Gobas (2004)

was developed for the RI/FS to estimate relationships among PCB concentrations in surface sediment, in the

water column, and in fish and shellfish tissue. The model has three environmental compartments: sediment,

porewater, and surface water; and five biological compartments: phytoplankton/algae; zooplankton; predatory

and filter-feeding benthic invertebrates; detritus-consuming benthic invertebrates; and fish.

Data for PCB concentrations in sediment, water, and tissue were combined with King County’s EFDC

hydrodynamic and contaminant fate and transport model to provide input parameters for the FWM (King

County 2010). The EFDC model used for the FWM was a predecessor to the more refined EFDC model used

in the STM (Section 5.1.2). The FWM was used to estimate PCB concentrations in water for the entire LDW

over different spatial scales and over various temporal scales that include all four seasons.

The FWM results were used to: 1) estimate risk-based threshold concentrations (RBTCs) for PCBs in sediment

for seafood consumption, and 2) estimate residual risks from PCBs that would remain after remediation under

the FS alternatives for use in the comparative analysis of alternatives (Sections 10.1.1 and 10.2.1).

The FWM was used only for PCBs. The relationship between concentrations in sediment and concentrations

in tissue for other seafood-consumption COCs was considered in the following ways:

For arsenic and cPAHs7, seafood consumption risks to humans were largely attributable to eating

clams. However, analyses performed during the RI showed relationships between concentrations in

clam tissue and in sediment were not statistically significant for either arsenic or cPAHs.

For dioxins/furans, available data indicated that dioxin/furan concentrations in most Puget Sound fish

and shellfish tissues would likely present unacceptable risk at the consumption rates used in the human

health risk assessment. Therefore, no more modeling or study was needed to support a conclusion that

the most appropriate dioxin/furan cleanup goals for LDW sediments would be background

concentrations.

5.2.2 Food Web Model Uncertainty

Input parameters and distributions for the FWM were based on literature-derived and site-specific

environmental data. Several analyses were performed to assess the sensitivity of the FWM to individual input

parameters in combination with the uncertainty in estimates of those parameters. These analyses are

summarized in the RI Report (LDWG 2010).

Key sources of uncertainty associated with use of the FWM for projecting post-cleanup residual risks are:

Only PCBs were modeled; thus, any estimates of present or future risks using the FWM underestimate

total risk by not including the other human-health seafood-consumption COCs.

The FWM was calibrated using data collected in the late 1990s through 2005. The FWM has not been

used with a different set of sediment and water concentrations to assess how accurately it can estimate

tissue concentrations outside the range to which the FWM was calibrated. It is unknown how

7 cPAHs consist of a subset of seven PAHs which EPA has classified as probable human carcinogens:

benz[a]anthracene, benzo[a]pyrene, benzo[b]fluoranthene, benzo[k]fluoranthene, chrysene,

dibenz(a,h)anthracene, and indeno(1,2,3-cd)pyrene.


22

predictive the model will be under lower sediment concentrations such as are projected after remedial

actions are completed.

There is uncertainty in the projected post-remedy sediment PCB concentrations that are a key input

parameter to the FWM. These post-remedy sediment PCB concentrations are based on the BCM,

which is subject to its own set of uncertainties, as described in Section 5.1.4.

EFDC model projections and best professional judgment were used to estimate a key input parameter

to the FWM: post-remedy water PCB concentrations. A sensitivity analysis showed that the water

PCB concentration becomes the most important factor for modeling tissue concentrations at the very

low projected post-cleanup sediment PCB concentrations.

5.3 Nature and Extent of Contamination

See Section 2.2 for a summary of LDW investigations prior to the RI/FS. LDWG completed Phase 1 of the RI

in 2003. This study compiled and analyzed pre-existing data, identifying areas of higher contamination to be

considered for early cleanup. LDWG collected extensive additional data during Phase 2 of the RI through

2009. These additional data, new data collected by other parties, and Phase 1 data were used to develop the RI

Report, which was completed in 2010. The RI study area initially extended to RM 7, but was later reduced to

the lower 5 miles in the FS because RI data showed very low levels of contamination upstream of RM 5.

The nature and extent of hazardous substance contamination was evaluated in the RI/FS based on the

concentration of contaminants in approximately 1,500 surface sediment samples (from within the top 10 cm

of the river bed), 900 subsurface sediment samples, 420 fish and shellfish tissue samples, 480 surface water

samples, 110 seep samples, and 90 porewater samples. Toxicity tests were performed on 76 surface sediment

samples. Approximately 900 of these samples were collected as part of Phase 2 of the RI in 2004 – 2006, and

an additional 47 samples were collected during the FS in 2009 – 2010. The rest of the dataset consisted of data

from samples collected during other investigations in the waterway (e.g., investigations conducted prior to the

start of the RI discussed in Section 2.2, and investigations conducted as part of the EAA cleanups) between

1990 and April 2010, which were incorporated into the RI and FS datasets. Because the RI report was

completed before the FS, the RI used a slightly different dataset than the FS. The RI baseline dataset included

data collected between 1990 and October 2006. The FS baseline dataset included additional data collected

after October 2006 and until April 2010. The expanded FS dataset filled data gaps but did not result in

significant changes to the CSM or assessment of human health or ecological risks. The results of these

investigations for sediments, fish and shellfish tissue, and surface water are described in the RI (LDWG 2010)

and FS (LDWG 2012a) reports and summarized below. Data shown in Table 1 are from the FS dataset.

5.3.1 Surface and Subsurface Sediments

Table 1 summarizes minimum and maximum detected concentrations, average concentrations, and detection

frequencies for PCBs, arsenic, cPAHs, and dioxins/furans in surface and subsurface sediments. Sediment

samples were analyzed for numerous contaminants during the RI/FS; these contaminants account for the

majority of human health risks from contamination in the LDW, as discussed in Section 7.1.

5.3.1.1 Surface Sediments

Based on RI data, PCBs are the most widespread contaminant in LDW surface sediment; they were detected at

94% of the locations where samples were analyzed for PCBs. The distribution of PCBs in LDW surface

sediment is shown in Figure 4.


23

Table 1. Statistical Summaries for Baseline Human Health COC Concentrations in Sediment

Data Type/Contaminant

Summary Statistics for Sediment in the LDW

(RM 0 to 5.0)

Total Number of

Sediment Samples in

FS Baseline Dataset

Minimum

Detect

Calculated

Mean

Maximum

Detect

Spatially-Weighted

Average

Concentration

(SWAC) Total

With

Detected

Values

Surface Sediment

PCBs (µg/kg dw) 2.2 1,136 220,000a 346 1,392

(1,390)a 1,309

Arsenic (mg/kg dw) 1.2 17 1,100 15.6 918 857

cPAHs (µg TEQ/kg dw)b 9.7 459 11,000 388 893 852

Dioxins/Furans (ng TEQ / kg dw)c 0.25 42 2,100 25.6 123 119

Subsurface Sediment

PCBs (µg/kg dw) 0.52 1,953 890,000 n/a 1,504 1131

Arsenic (mg/kg dw) 1.2 29 2,000 n/a 531 453

cPAHs (µg TEQ/kg dw)b 1.2 373 7,000 n/a 542 449

Dioxins/Furans (ng TEQ / kg dw)c 0.15 17 194 n/a 64 64

Source: FS baseline surface and subsurface sediment dataset dated April 28, 2010 (surface) and May 14, 2010 (subsurface). This summary is based on data used for the Feasibility Study (see Section 5.3); however, for the human health and ecological risk assessments the RI dataset was used..

a. This table excludes two PCB samples, both collected at the inlet at RM 2.2. They were considered anomalous samples and statistical outliers and were not included in calculated mean and SWAC; their detected concentrations were 230,000 and 2,900,000 µg/kg. If the outliers were included, the mean would be 3,400 μg/kg dw and the SWAC would be 1,300 μg/kg dw.

b. The cPAH TEQs were calculated using compound-specific potency equivalency factors. c. The dioxin/furan TEQs were calculated using World Health Organization’s mammalian toxic equivalent factors.


24

Figure 4. PCB Distribution in Surface Sediment


25

Areas where COCs were present at concentrations toxic to benthic invertebrates were identified in the RI/FS

based on the Washington State Sediment Management Standards (SMS).

The SMS were revised on September 1, 2013, after issuance of the Proposed Plan; therefore some terminology

used in the Proposed Plan has been changed for this ROD. Among other changes the revised SMS changes the

term for the criteria for protection of marine benthic invertebrates to “sediment cleanup objectives" or SCO,

formerly called “sediment quality standards” (SQS). This is a terminology change only; the benthic protection

criteria themselves are unchanged. See "What are the Sediment Management Standards?" on page 26 for a

summary of the revised SMS.

This ROD uses the terms "benthic SCO" and "benthic CSL" to refer to the criteria to protect benthic

invertebrates. The more general terms “SCO” and “CSL” refer to the overall standards which include

standards for the protection of human health (HH) and for the protection of higher trophic level species

(HTLS) that have been referred to more simply as wildlife in this ROD. On occasion, the terms HH SCO and

HTLS SCO are used.

In the SMS, the numerical sediment cleanup objectives (benthic SCOs) are contaminant concentrations below

which no adverse effects on benthic invertebrate organisms are expected. The SMS also establishes cleanup

screening levels (benthic CSLs), higher levels for the same contaminants at which minor effects are expected.

These benthic SCOs and benthic CSLs, together, are called chemical criteria. The SMS regulations also allow

use of alternate criteria — site-specific biological-effects criteria that are based on toxicity testing or benthic

abundance data — to determine whether a location does or does not meet (i.e., passes or fails) the benthic SCO

or benthic CSL; these are called biological criteria.

Surface sediment samples from 76 locations were tested for both contaminant concentrations and biological

effects.8 Figure 5 shows areas where neither benthic SCOs nor CSLs were exceeded (areas identified as

“pass”), areas where only SCOs but not CSLs were exceeded, and areas where CSLs were exceeded.

Forty-one hazardous substances were designated as COCs for benthic invertebrates because they were detected

in LDW sediment at concentrations that exceed the benthic chemical SCOs and. In surface sediment, PCB

concentrations exceeded the benthic SCO more frequently than those of any other COC, followed by BEHP,

then butyl benzyl phthalate. The locations where these exceedances occurred and extent of these exceedances

varied by COC. Section 7.2.5 provides summary information on surface sediment contamination for these 41

COCs, including minimum and maximum detected concentrations, average concentrations, detection

frequencies, and exceedances of benthic SCOs and benthic CSLs.

In addition to contaminant concentrations, several other parameters were measured during the RI. The percent

fines (sum of clay and silt fractions) in LDW surface sediments ranged from 13 to 87% with an average of

approximately 53%, and the LDW-wide average total organic carbon (TOC) content is 2%. Sediment grain

size and TOC content influence the quality of habitat for benthic invertebrates and other organisms. Grain size

is also important in determining whether sediments will erode from or be deposited in the LDW. TOC

influences the bioavailability of some organic contaminants. Because of this, many organic contaminants are

“normalized” to TOC in the SMS.

8 Sample locations where benthic SCO or benthic CSL chemical criteria were exceeded but benthic SCO or CSL

biological criteria were not exceeded were designated as not exceeding the benthic SCO or CSL—that is, the

determination of whether criteria are exceeded was based on biological criteria not chemical criteria. It is important to

note that risks to human health or to fish and wildlife coming into contact with sediment or eating fish and shellfish

that live in the waterway are not addressed by either the SMS numerical chemical criteria or the SMS biological

criteria. Those risks are addressed separately as described in this ROD.


26

What are the Sediment Management Standards (SMS)?

The SMS are State standards designed to reduce and ultimately eliminate adverse effects on biological resources and significant health

threats to humans from surface sediment contamination.

The SMS uses two tiers to establish a sediment cleanup level. The sediment cleanup level starts at the lower tier and may be adjusted no

higher than the upper tier (provided certain factors are met).

Tier 1) sediment cleanup objectives (SCOs), the level that is the environmental goal for establishing sediment cleanup levels; and

Tier 2) cleanup screening levels (CSLs), the level used to identify cleanup sites and the maximum level for establishing sediment

cleanup levels.

The sediment cleanup level for each COC is initially set at the SCO and is allowed to be adjusted upward (but cannot exceed the CSL) if it is

not technically possible to achieve the SCO or if achieving the SCO will result in a net adverse environmental impact.

Each tier consists of:numerical chemical concentration criteria (chemical criteria) and biological effects criteria for protection of benthic

invertebrates, risk-based standards for protection of human health, narrative standards for protection of other ecological receptors, and

provisions for use of background concentrations. The chemical criteria for protection of marine benthic invertebrates are based on

relationships between sediment contaminant concentrations and toxicity to adverse effects on benthic invertebrates (reduced population

size or laboratory toxicity tests showing mortality, reduced growth, or impaired reproduction) using several hundred samples from the Puget

Sound area. The biological-effects criteria allow for site-specific testing for toxicity or for abundance.

The SCO for each COC is established as the highest (least stringent) of risk-based concentrations, natural background concentrations, and

practical quantitation limits (PQL). The risk-based criteria are the lowest (most stringent) of:

concentrations protective of human health (based on an excess cancer risk of 1 x 10-6 for individual carcinogens, 1 x 10

-5 for

carcinogens culmulatively, or a noncancer Hazard Quotient [HQ] or Hazard Index [HI] of 1.0) (HH SCO)

concentrations showing no adverse effects to the benthic community (benthic SCO), specified as chemical criteria and biological

criteria in the SMS, and

concentrations resulting in no adverse impacts to higher trophic level species (HTLS SCO).

The CSL for each COC is established as the highest (least stringent) of risk -based concentrations, regional background, and PQL. The risk-

based criteria are the lowest (most stringent) of:

concentrations protective of human health (based on 1 x 10-5 excess cancer risk for individual carcinogens and carcinogens

cumulatively, or a noncancer HQ or HI of 1.0) (HH CSL),

concentrations showing minor adverse effects to the benthic community (benthic CSL), specified as numerical chemical and

biological criteria in the SMS, and

concentrations resulting in no adverse impacts to HTLS (HTLS CSL).


27

Figure 5. SMS Status in Surface Sediment


28

5.3.1.2 Subsurface Sediments

Detected contaminant concentrations were above the benthic SCO for one or more of the SMS

contaminants in 49% of the subsurface sediment samples. The average thickness of subsurface sediments

with COC concentrations greater than the benthic SCO is 4 ft. Of the COCs detected in subsurface

sediments at concentrations above the benthic SCO, PCBs were detected most frequently (48% of the

samples), followed by BEHP (25% of the samples). Table 1 summarizes sediment contaminant

concentration data (both surface and subsurface) for the human health COCs: PCBs, arsenic, cPAHs, and

dioxins/furans. (Selection of human health COCs is discussed in Section 7.1.1).

5.3.2 Fish, Shellfish, and Benthic Invertebrate Tissue

The ranges of contaminant concentrations discovered during the RI/FS for all types of organisms and

tissue types (in some cases whole organisms and in other cases portions of the organisms), and the general

trends in tissue concentrations observed are summarized as follows:

PCBs – Detected in almost all samples, ranging from 6.9 to 18,400 micrograms per kilogram wet

weight (µg/kg ww). Mean PCB concentrations were highest for Dungeness crab hepatopancreas

(“crab butter”) and whole-body English sole, followed by whole-body shiner surfperch. Clam and

crab edible meat had much lower mean PCB concentrations, with mean PCB concentrations

being lowest for mussels.

Inorganic arsenic – Detected in almost all samples, ranging from 0.003 milligrams per kilogram

wet weight (mg/kg ww) to 11.3 mg/kg ww. Inorganic arsenic is the most toxic form for humans

and wildlife. Total arsenic was measured in sediments and water and inorganic arsenic was

measured in fish and shellfish tissue. Eastern softshell clams (the most abundant clam species

found in the LDW) had the highest average concentrations of inorganic arsenic, approximately

3 mg/kg ww. Inorganic arsenic concentrations in clams in the LDW were one to three orders of

magnitude greater than inorganic arsenic concentrations found in other organisms.

Other COCs

Concentrations of cPAHs were highest in clam, mussel, and benthic invertebrate tissue.

Phthalates were frequently detected in clams and benthic invertebrates. Most other organic

hazardous substances were infrequently detected.

Sampling for dioxins/furans in tissue was not included as part of the RI because data that

were already available indicated that dioxin/furan concentrations in most Puget Sound fish

and shellfish tissues would present unacceptable risk at the consumption rates used in the

human health risk assessment. No further data were needed for the human health risk

assessment, which assumed unacceptable risks due to dioxins/furans. Separate from the RI,

data were gathered from a small number of dioxin/furan samples of skin-off English sole

fillets collected near Kellogg Island by Ecology in May 2007; however, this was after

completion of the human health risk assessment. See Section 7.1 for further discussion.

Table 2 summarizes PCB, inorganic arsenic, cPAH, and dioxin/furan tissue concentrations in the LDW

fish and shellfish collected and analyzed in the RI.

Tissue PCB concentrations in benthic invertebrates were often higher in areas with higher sediment PCB

concentrations. A similar pattern was observed for some species of fish (shiner surfperch, and staghorn

sculpin) indicating that they may have smaller foraging ranges; in others (English sole and crabs), tissue


29

concentrations did not show a clear relationship to sediment concentrations, indicating they may have

larger foraging ranges. Clams had the highest inorganic arsenic and cPAH tissue concentrations, but no

strong relationship was seen between sediment concentrations and clam tissue concentrations of arsenic

and cPAHs.

Table 2. Summary of Selected Baseline Human Health COC Concentrations in Fish and Shellfish Tissue

a

Contaminant and Tissue Type

Detection

Frequency

Concentration

Minimum Mean Maximum

PCBs (µg/kg ww)

English sole (fillet with skin) 26/26 170 860 2,010

Shiner surfperch (whole body) 78/78 200b 1,300 18,400b

Dungeness crab (edible meat) 14/17 15 130 300

Dungeness crab (whole bodyc) 16/16 97 890 1,900

Clams 20/20 15b 130 580b

Inorganic Arsenic (mg/kg ww)

English sole (fillet with skin) 6/7 0.003 0.004 0.006

Shiner surfperch (whole body) 8/8 0.020 0.070 0.160

Dungeness crab (edible meat) 2/2 0.010 b 0.010 0.010

Dungeness crab (whole bodyc) 2/2 0.022b 0.029 0.035

Clams 23/23 0.132 2.72 11.3

cPAHs (µg TEQ/kg ww)

English sole (fillet with skin) 4/7 0.37b 0.35 0.53

Shiner surfperch (whole body) 24/27 0.37b 3.1 2.2

Dungeness crab (edible meat) 6/9 0.54b 3.7b 0.84b

Dungeness crab (whole bodyc) 7/9 0.60 b 2.6 2.4 b

Clams 14/14 6.8 15 44

Dioxins/Furans (ng TEQ/kg ww)d

English sole (fillet without skin) 6/6 0.26 0.30 0.35

Note: Additional fish and shellfish species were collected during the RI. All fish and shellfish COC data are presented in the RI. a. Section 7.1.1 describes the selection of human health COCs. b. These data points are analytically estimated values. c. Whole body Dungeness crab concentrations were estimated based on results from edible meat and hepatopancreas samples. d. The dioxin/furan data are from samples collected in a small portion of the LDW as part of a 2007 Ecology study, and were not used in

the LDW risk assessments.

5.3.3 Surface Water

The water column, along with contaminated sediments, is an important pathway for COCs to reach

benthic organisms, fish, and shellfish. LDW surface water was collected and analyzed by King County for

metals, semi-volatile organic compounds, and PCBs in 1996 and 1997 and for PCBs in 2005. For the

human health COCs, PCB concentrations in 2005 LDW unfiltered surface water samples ranged from

0.13 to 3.2 nanograms per liter (ng/L), with the lowest concentrations detected during periods when flows

were highest. Dissolved arsenic concentrations ranged from 0.18 to 1.5 micrograms per liter (µg/L).

Detected surface-water PAH concentrations (collected using semipermeable membrane devices) ranged

from 0.0027 ng/L for dibenzo(a,h)anthracene to 0.35 ng/L for benzo(a)anthracene. Dioxins/furans were

not measured in surface water due to the difficulty in detecting these contaminants in whole water

samples. EPA determined that more water quality sampling during the RI would not have affected the


30

analysis of human health or ecological risks, or have influenced the development of alternatives for the

In-waterway Portion of the Site.

5.3.4 Background COC Concentrations

Documenting background concentrations (concentrations at locations away from the Site and away from

other sources of contamination) is important in the process of identifying cleanup goals. This section

describes how contaminant concentrations in sediments and in fish and shellfish in non-urban areas of

Puget Sound were used to estimate background conditions.

5.3.4.1 Sediment Background Concentrations

Under CERCLA and MTCA/SMS, when risk-based threshold concentrations (RBTCs, see Section 8) are

below background, background concentrations are used as cleanup levels. The State of Washington

adopted its revised SMS in September 2013. The 2013 SMS sets two background levels. The first is the

Sediment Cleanup Objective (SCO), which sets the background concentration at the natural background

level. Natural background under MTCA is defined for sediments in WAC 173-204-505 (and for all other

media in WAC 173-340-200) as “the concentrations of a hazardous substance consistently present in the

environment that has not been influenced by localized human activities.” Thus, under MTCA, a natural

background concentration can be defined for human-made compounds even though they may not occur

naturally. For example, PCBs (human-made compounds) can be picked up and carried by the winds and

then deposited into an alpine lake that has not been locally influenced by human activities, and the

concentration of PCBs that is then consistently present in that lake is the natural background level. A

second, new, background level is defined under the 2013 SMS rule at the Cleanup Screening Level

(CSL), where the background concentration may be set at "regional background." Regional background

concentrations for this Site have not been established. See Section 8.2.2.1 for further discussion of the

2013 SMS rule and the possible future use of regional background. The MTCA/SMS approach is

consistent with EPA guidance (EPA 2002a), which calls for the use of natural or anthropogenic

background as appropriate for the circumstances at a particular site.

To characterize sediment natural background COC concentrations, data from a 2008 study (USACE et al.

2009) of sediment contaminant concentrations in non-urban areas in Puget Sound were used. Sediment

samples were collected at locations that are away from populated and industrial areas and known

contaminated sites. Summary statistics were then calculated for each of the four human health COCs.

Table 3 summarizes these data.

Table 3. Summary of PCB, Arsenic, cPAH, and Dioxin/Furan Data for Natural Background Concentrations in Sediment

Human Health COC

Detection

Frequency

Concentration

Minimum Maximum Mean Median

90th

Percentile

95th Percentile Upper

Confidence Limit on

the Mean (UCL95)a

PCBs (μg/kg dw)b 70/70 0.01 11 1.2 0.6 2.7 2

Arsenic (mg/kg dw) 70/70 1.1 21 6.5 5.9 11 7

cPAHs (μg TEQ/kg dw) 61/70 1.3 58 7.1 4.5 15 9

Dioxins/Furans (ng TEQ/kg dw) 70/70 0.2 12 1.4 1.0 2.2 2

a. These values are rounded to one significant figure. b. Only PCB congener data from the EPA (2009a) study were used, as there were few detected values in the Aroclor data.


31

5.3.4.2 Fish and Shellfish Tissue Background COC Concentrations

To define natural background COC concentrations, a dataset of COC concentrations in fish and shellfish

tissue samples collected between 1991 and 2009 from non-urban areas in Puget Sound, away from

populated and industrial areas and known contaminated sites, was compiled for each of the four human

health COCs (PCBs, inorganic arsenic, cPAHs, and dioxins/furans). These non-urban Puget Sound fish

and shellfish tissue data, shown in Table 4, were used to set target tissue concentrations (Section 8.2.3,

Table 21) when RBTCs were below background.

Table 4. Summary of PCB, Arsenic, cPAH, and Dioxin/Furan Data for Natural Background Concentrations in Fish and Shellfish Tissue

Species

Natural Background Fish and Shellfish Tissue Data

Detected

Samples /

Total Samples

Range of

Detected

Concentra-

tions Mean

95th Percentile

Upper Confidence

Limit on the Mean

(UCL95)

PCBs (μg/kg ww)

English sole, rock sole (fillet) 158 / 238 1.3 – 75.4 11 12

Dungeness crab (edible meat) 17 / 17 0.43 – 1.9 0.87 1.1

Dungeness crab (whole body) 15 / 15 3.0 – 16 7.1 9.1

Butter clam, geoduck, horse clam, littleneck clam (whole body)

24 / 70 0.09 – 1.4 0.3 0.42

Inorganic arsenic (mg/kg ww)

Eastern softshell clams (whole body)a,b 6 / 0 0.047 / 0.112 0.064 0.09

cPAH TEQ (μg/kg ww)

Butter clam, geoduck, littleneck clam (whole body)a 3 / 11 0.069 – 0.17 0.088 0.12

Dioxin/furan TEQ (ng/kg ww)

Starry flounder, rock sole (whole body)c 7 / 7 0.17 – 0.92 0.28 0.35

Dungeness crab (edible meat) 27 / 27 0.027 – 1.4 0.57 0.53

Dungeness crab (whole body) 25 / 25 0.089 – 5.1 0.81 2.0

Butter clam, geoduck, horse clam, littleneck clam (whole body)

43 / 43 0.011 – 1.6 0.34d 0.71

a. Only clams are shown for inorganic arsenic and cPAH TEQ because most of the risk associated with these COCs was due to consumption of clams.

b. Only clams collected from Dungeness Spit were selected by EPA for this category, as these were the only ones in the dataset likely unaffected by the atmospheric deposition of arsenic from the former Tacoma ASARCO smelter.

c. There were insufficient data to derive a background value for pelagic fish (e.g., perch) for total PCBs, cPAHs, and dioxins/furans; there were insufficient data for benthic fish (e.g., English sole) fillets for dioxins/furans.

d. This is a nonparametric mean, as there was no discernible distribution according to ProUCL v. 4.1.

Background COC concentrations for non-urban Puget Sound fish and shellfish tissue are much more

uncertain than background concentrations for sediment. The dataset is comprised of data from various

studies representing different sampling and analysis methods. It also contains widely differing numbers of

samples for the various COCs and tissue types, depending on data availability and data quality

considerations. No tissue data were collected upstream of the Site because the river conditions transition

from marine and estuarine to a freshwater environment, with different fish and shellfish species.


32

5.3.5 Sediment COC Concentrations from Upstream of the LDW Study Area

This section describes the information used in the BCM to estimate the concentrations of COCs in

upstream sediments that deposit in the LDW.

Several datasets with sediment COC concentrations from upstream locations were evaluated for use in

estimating COC concentrations in suspended sediments entering the LDW from the Green/Duwamish

River. Because of the large volume of suspended sediments entering the LDW from the Green/Duwamish

River, these data were important input parameters for BCM-predicted estimates of future COC

concentrations in the LDW after implementation of the cleanup alternatives evaluated in the FS (see

Section 9.4).

Datasets that were included:

COC concentrations in Green/Duwamish River surface sediments and suspended sediments

immediately upstream of the Site from two 2008 Ecology studies. Surface sediments in the

Green/Duwamish River are generally much coarser than those found in the LDW. In order to

match the grain size of sediments that deposit in the LDW, only surface sediments with greater

than 30% fines were included.

Sediment core data collected from the LDW Upper Turning Basin from 1991 to 2009 by USACE

for maintenance dredging. The LDW Upper Turning Basin is a sink for sediments entering the

LDW from upstream. These data provide an indicator of suspended sediments settling in the

upper reach of the LDW.

For the four human health COCs, Table 5 shows the low, high, and mid-range estimates of upstream

suspended sediment COC concentrations selected for use in the BCM. Each sampling technique may

over- or underestimate the COC concentrations in sediments entering the LDW. In addition, all of the

datasets used in this evaluation were small (e.g., datasets used for the upstream input parameters ranged

from 6 to 31 samples). All these factors make the estimates of incoming sediment COC concentrations

more uncertain than background sediment concentrations. Best professional judgment was used to select

mid-range values used as upstream input values for the BCM, and also to select high and low values used

for sensitivity analysis.

Table 5. Estimates of Upstream Suspended Sediment Concentrations of PCBs, Arsenic, cPAHs, and Dioxins/Furans Used in the LDW Bed Composition Model

Human Health

COCs

BCM Parameters

Basis for BCM Upstream Input Values and Sensitivity Analysis Valuesa Input Low High

PCBs

(μg/kg dw)

35 5 80 Input: Upper Turning Basin subsurface sediment data, mean (36 rounded to 35). Low:

Ecology upstream Green/Duwamish River surface sediment samples containing fines > 30%,

mean. High: TSS-normalized King County (whole-water), UCL 95 (82 rounded to 80).

Arsenic

(mg/kg dw)

9 7 10 Input: Ecology upstream Green/Duwamish River surface sediment samples containing fines

>30%, mean. Low: Upper Turning Basin subsurface sediment data, mean. High: UCL 95,

Ecology upstream Green/Duwamish River surface sediment samples containing fines >30%.

Carcinogenic

PAH

(μg TEQ/kg dw)

70 40 270 Input: Upper Turning Basin subsurface sediment data, mean (73 rounded to 70). Low:

Ecology upstream Green/Duwamish River surface sediment samples containing fines >30%,

mean (37 rounded to 40). High: TSS-normalized King County (whole-water), UCL 95 (269

rounded to 270).


33

Human Health

COCs

BCM Parameters

Basis for BCM Upstream Input Values and Sensitivity Analysis Valuesa Input Low High

Dioxins/Furans

(ng TEQ/kg dw)

4 2 8 Input: Ecology upstream Green/Duwamish River surface sediment and upstream suspended

sediment data, midpoint between means. Low: Ecology upstream Green/Duwamish River

surface sediment samples containing fines > 30%, mean. High: Ecology upstream

suspended sediment data, midpoint between mean and UCL 95.

a. Upstream BCM parameter values were revised using updated datasets and statistics reflective of current conditions (i.e., material entering the LDW from the Green/Duwamish River). The four primary datasets used for BCM parameterization are:

Ecology’s 2008 upstream bed sediment chemistry data: This dataset was screened to exclude samples with ≤30% fines in consideration of the systematic differences in grain size distributions between upstream (e.g., mid-channel) data and average conditions in the LDW.

Total Suspended Solids (TSS)-normalized King County data: King County surface water data were normalized to solid fractions by dividing by the TSS in the individual sample.

Ecology 2008 centrifuged suspended solids data: The Ecology samples are representative of sediments suspended mid-channel in the Green/Duwamish River that enter the LDW.

Upper-reach USACE Dredged Material Management Program (DMMP) core data (RM 4.3 to RM 4.75): This dataset is representative of Green/Duwamish River suspended material that settles in the upper section of the LDW.

6 Current and Potential Future Land and Waterway Use This section summarizes the current and reasonably anticipated future use of the waterway and

surrounding watershed. This information forms the basis for the exposure assessment assumptions listed

and discussed in Section 7.1.

6.1 Land Use

The LDW and surrounding area is Seattle’s primary industrial corridor. Industries currently operating

along the Duwamish include marine construction, boat manufacturing and repair, marinas, cement

manufacturing, cargo handling and storage, paper and metals fabrication, food processing, airplane parts

manufacturing, and a municipal airport. However, the Duwamish estuary subwatershed (extending from

RM 11 to Elliott Bay) of the Green/Duwamish watershed has more residential land use (36%) than

industrial and commercial land use combined (29% combined; 18% and 11%, respectively). Eighteen

percent of the subwatershed is used for right-of-way areas (including roads and highways); and 17% is

open/undeveloped land and parks.

Residential areas near and on the LDW include the neighborhoods of South Park and Georgetown. These

neighborhoods support a mixture of residential, recreational, commercial, and industrial uses. EPA and

Ecology have identified environmental justice concerns in the South Park and Georgetown neighborhoods

in accordance with Executive Order 12898, Federal Actions to Address Environmental Justice in Minority

Populations and Low-Income Populations. As noted in EPA's Environmental Justice Analysis (EPA

2013a), incomes in South Seattle where these neighborhoods are located are approximately 50% lower

and percentages of minority populations are significantly higher than in the City of Seattle (the population

of the City of Seattle is approximately 30% minority, compared to the LDW corridor which is

approximately 50% minority). This area also has higher rates of asthma hospitalizations and higher rates

of other chronic diseases such as diabetes than other Seattle neighborhoods and King County as a whole.

6.2 Waterway Use

The LDW supports major shipping activities for containerized and bulk cargo. Approximately 40 berthing

areas are located along the LDW. Four marinas permit live-aboard vessels. The LDW is also used for


34

commercial salmon fishing and various recreational activities such as boating, kayaking, fishing, and

beach recreation. The LDW serves as a habitat for fish, shellfish, and wildlife, as described in Section 7.2.

Designated uses for the LDW can be found under WAC 173-201A-610 and 612 (Table 612). The LDW

is considered marine water under the state’s water quality standards regulation because it meets the

salinity threshold (vertically averaged maximum daily salinity of 1 part per thousand qualifies as marine)

described in WAC 173-201A-260(3)(e). Salinity measurements show tidal condition exists beyond the

turning basin. Although the Lower Duwamish Waterway is not specifically mentioned in Table 612, the

LDW is considered a continuation of Elliott Bay for the purposes of applying marine criteria.

Specifically, the same marine designated uses and standards apply east of a line between Pier 91 and

Duwamish Head as long as the salinity threshold is met. The designated uses for the LDW include:

Aquatic life uses:

o Excellent: Excellent quality salmonid and other fish migration, rearing, and spawning;

clam, oyster, and mussel rearing and spawning; crustaceans, and other shellfish (crabs,

shrimp, crayfish, scallops, etc.) rearing and spawning

Shellfish harvest:

o Shellfish (clam, oyster, and mussel) harvesting

Recreational uses:

o Primary contact recreation

Miscellaneous uses:

o Wildlife habitat

o Harvesting (Salmonid and other fish harvesting, and crustacean and other shellfish (crabs,

shrimp, scallops, etc.)

o Commerce and navigation

o Boating

o Aesthetic values



manages aquatic resources north of the Spokane Street Bridge, just north of the LDW study area.

The Washington State Department of Health (WDOH) currently maintains a seafood consumption

advisory recommending no consumption of resident fish and shellfish from the LDW. As discussed in

Section 7.1, the limited amount of time salmon spend in the LDW results in low site-related salmon body

burdens of bioaccumulative contaminants. Consequently, salmon are not included in the LDW fish

advisory, although they are included in a South Puget Sound advisory which recommends eating no more

than one meal per week to one meal per month depending on the species. The WDOH maintains a web

site and provides publications and other educational forums that cover healthy eating and seafood

consumption. In addition, the seafood consumption advisories are posted on signs at public access

locations within the LDW. More information can be found at http://www.doh.wa.gov/fish. In spite of the

seafood consumption advisory, fishers report regularly catching and consuming LDW resident fish and

shellfish (LDWG 2014b).

http://www.doh.wa.gov/fish


35

The LDW shoreline is zoned predominantly for recreation, conservancy preservation, and urban industrial

land use. A small area on the west side of the LDW between RM 3.1 and RM 3.3 is zoned for single-

family residential use. Several public parks and publicly accessible shoreline areas exist within the LDW,

and there are plans to create additional recreational and habitat opportunities in the LDW corridor. Four

marinas and two public parks (Terminal 107/Herring’s House and Duwamish Waterway Park) are located

along the LDW, and several other access points allow the public to enter the LDW for recreational

purposes. A non-Federally recognized Tribe, the Duwamish Tribe, uses parks along the LDW for cultural

gatherings and canoe launching. A human access survey conducted along the LDW shoreline as part of

the RI survey identified the following uses: launching and hauling out hand-powered boats or motorboats,

walking, fishing, swimming, and picnicking (Figure 6). Figure 6 also identifies beach play areas

(intertidal areas accessible from shore) and potential clamming areas (areas where clams are present) used

in the human health risk assessment.

Although habitat and recreational use, as well as Tribal fishing and shellfishing, may increase at some

point in the future, the reasonably anticipated future use of the waterway and surrounding area are

anticipated to remain similar to current use.


36

Figure 6. LDW Areas with Parks and Habitat Restoration, Beach Play Activities, and Potential Clamming.


37

7 Summary of Site Risks Baseline human health and ecological risk assessments (HHRA and ERA) were conducted during the RI

to determine potential pathways by which people (human receptors) or animals (ecological receptors)

could be exposed to contamination in seafood, sediments, or water, the amount of contamination

receptors of concern may be exposed to, and the toxicity of those contaminants if no action were taken to

address contamination at the Site. These assessments provide the basis for taking action and identify the

contaminants and exposure pathways that need to be addressed by the remedial action. Multiple exposure

pathways by which humans or animals could be exposed to contaminants in the In-waterway Portion of

the Site (the "waterway" or LDW) were evaluated. Figure 7 shows the human health conceptual site

model. Esitmates of risks remaining after cleanup (Section 10) can be compared against the baseline

assessments to determine the amount of improvement that results from a remedial action.

Figure 7. Conceptual Model for Baseline Human Health Risk Assessment

7.1 Human Health Risks

In conducting a human health risk assessment, EPA evaluates the potential for noncancer health effects

such as immunological, reproductive, developmental, or nervous system disorders, and the potential for


38

increased cancer risk. Different methods are used to estimate noncancer health effects than cancer risks.

EPA estimates human health risks using the following process:

In Step 1, EPA gathers and analyzes data on the concentrations of contaminants found at a site to

identify the contaminants that will be the focus of the risk assessment (called contaminants of

potential concern, or COPCs).

In Step 2, EPA considers the amount of a contaminant that may enter a person’s body at the Site

(i.e., the dose). EPA considers different scenarios in which people might be exposed to the

COPCs (e.g., playing at the beach, fishing, or eating seafood). For each scenario, the

concentration of COPCs that people might contact (e.g., in sediment or in fish tissue), and other

information such as frequency and duration of fish consumption are used to compute the dose for

that scenario. The ways that people might be exposed (pathways) are also considered and there

may be more than one pathway in each scenario.

In Step 3, EPA considers information on cancer toxicity (i.e., cancer slope factors [SFs]) or

noncancer toxicity (i.e., reference doses or RfDs) of each COPC, using scientific studies on the

effects of these contaminants on people or animals. Toxicity information is not specific to this

site, it is from sources that have studied the contaminants to find out how toxic they are no matter

where a person gets exposed to them.

In Step 4, EPA uses the information from the three previous steps to calculate site-related cancer

or noncancer risks. A site-related cancer risk is an extra or excess risk, in addition to other cancer

risks people are exposed to. EPA evaluates whether these risks are great enough to potentially

cause health problems for people at or near the Superfund site. The assumptions EPA uses in

calculating risks result in estimates that err on the side of protecting the public. The likelihood of

any kind of cancer resulting from exposure to contaminants at a Superfund site is generally

expressed as a probability; for example, a "1 in 1,000,000 chance" of developing cancer over the

course of a lifetime (in exponential form, 10-6

). In other words, for every million people that

could be exposed, one extra cancer may occur as a result of exposure to site contaminants. EPA

sums the cancer risks for each contaminant included in a particular exposure scenario to develop

a total risk value for that exposure scenario (e.g. seafood consumption risks posed by PCBs,

arsenic, etc.). EPA is generally concerned when site related risks exceed a range of 1 in 10,000 or

10-4

to 1 in 1,000,000 or 10-6

. For noncancer health effects, EPA also calculates a noncancer

"hazard quotient" (HQ). The hazard quotient is the dose of a contaminant a person might be

exposed to, divided by that contaminant’s RfD. The RfD is the amount of a contaminant a person

could take in over the course of a lifetime without expectation of adverse health effects. For

individual contaminants, if the ratio is 1 or less (i.e., the dose received is equal to or less than the

RfD), then no noncancer effects are expected. EPA is also concerned about the effects of multiple

contaminants with the same noncarcinogenic toxic effect and evaluates this by summing the HQs

for each relevant chemical. This sum is called the “hazard index” (HI), and is only of concern if

the HI exceeds 1.

In addition to the human health risk assessment, an uncertainty analysis summarizes the assumptions used

to compute risks, whether or not these assumptions lead to over- or underestimation of risk, and where

possible, quantifies the magnitude of uncertainty.


39

7.1.1 Identification of Contaminants of Potential Concern

In the LDW HHRA, COPCs are contaminants that were detected in sediments, water9, fish, and shellfish

at concentrations that exceeded risk-based screening criteria and were carried through the risk assessment

process. Using the dataset defined for the HHRA,10

a contaminant was designated a COPC if the

maximum detected concentration in sediment or tissue exceeded a health-protective risk-based screening

concentration (screening concentration).11

A contaminant was also designated a COPC if it was not

detected at any concentration above its analytical reporting limit (RL) but more than 10% of the samples

had RLs that exceeded the screening concentration. In such cases, one-half the maximum RL was the

value used to compute risks. The risk-based screening identified the following COPCs, by exposure

pathway: 59 COPCs for seafood consumption pathways, 20 COPCs for netfishing, and 28 COPCs for

beach play and clamming direct contact pathways. COPCs that were not detected in either sediment or

tissue were still included if they had RLs above the screening criteria; however, they were evaluated only

in the uncertainty analysis. Consideration of whether the contaminant had been used by industries

historically or currently present at the site was an important factor in evaluating whether or not a

contaminant with RL above screening criteria was actually present and should be included as a COPC.

The HHRA estimated risks for all COPCs, in order to identify and prioritize those contaminants that were

estimated to pose an unacceptable risk and therefore needed to be addressed in the FS. EPA and Ecology

regulations and guidance were then used to narrow the list of COPCs to a shorter list of four

“contaminants of concern” or COCs, for which risk-based threshold concentrations (RBTCs) were

developed for the FS: PCBs, arsenic, cPAHs and dioxins/furans (see text box, “Contaminants of Concern

for Human Health in LDW” below). Section 7.1.4 provides information on estimated risks for all COPCs

and selection of COCs.

9 COPCs and risks due to contact with water were evaluated using a 1999 King County analysis for the LDW and

Elliott Bay, see Section 7.1.2.1. 10

The HHRA and ERA datasets are slightly different than the FS datasets summarized in Section 5.3, because

additional data were collected after completion of calculations for these reports. 11

For contaminants with noncancer toxicity, a lower HQ criterion 0.1 was used as a health-protective screening

level to ensure that no contaminants were missed for further consideration.

Contaminants of Concern for Human Health in the LDW

Many hazardous substances are found in LDW sediments, fish, and shellfish. Most human health risk comes from these four:

PCBs are manmade chemicals that were banned in the late 1970s. PCBs were widely used in coolants and oils, paints, caulking,

and building material. PCBs stay in the environment for a long time and can build up in fish and shellfish. Children exposed to

PCBs may develop learning and behavior problems later in life. PCBs are known to have immune and reproductive system effects,

and may cause cancer in people who have been exposed to them over a long time.

Arsenic is associated with industrial uses like lumber treatment and watercraft repair. It is naturally present at low levels in Puget

Sound area rock and soil, and industrial activities have spread additional arsenic over much of the Puget Sound Region. Long-term

exposure to toxic forms of arsenic may cause nervous and cardiovascular toxicity, liver and kidney disorders, and several types of

cancer including skin and bladder.

PAHs are formed during the burning of substances such as coal, oil, gas, wood, garbage, and tobacco and during the charbroiling

of meat. Historical industrial activities are a known source of PAHs, as well as creosote treated timber. Long periods of breathing,

eating, or having skin contact with high levels of some of the PAHs may increase a person’s risk of cancer.

Dioxins/furans are by-products of burning (in either natural or industrial settings), chemical manufacturing, and metal processing.

Historically, dioxins/furans were byproducts of pentachlorophenol (used in wood treating), pesticide, and PCB production. Dioxins

last a long time in the environment and, like PCBs, can build up in fish and fatty foods. Specific toxic effects related to dioxins

include reproductive problems, fetal development or early childhood problems, immune system damage, and cancer.


40

7.1.2 Exposure Assessment

The exposure scenarios considered by EPA are shown in Table 6. Only the scenarios listed in Table 6 as

having "numeric" analysis were selected for evaluation in the HHRA; those listed as “qualitative” were

excluded from evaluation in the HHRA. The rationale for selecting or excluding each pathway is shown

in Table 6.

Consistent with EPA’s Human Health Risk Assessment guidance (EPA 1989) and Executive Order

12898, "reasonable maximum exposure" (RME) scenarios for several populations of interest were used in

the HHRA. The HHRA also considered three other scenario types: “central tendency” (CT), “upper-

bound” (UB), and “informational” (I). RME estimates are used by EPA for decision-making; they

represent the highest exposures that are reasonably expected to occur at a site, and were calculated for all

exposure scenarios to avoid underestimating risks. CT exposures were not used in decision-making

because they may underestimate exposure for a substantial number of individuals (EPA 1989). CT and

UB estimates are used to evaluate the range of uncertainty in risk estimates. “I” risk estimates provide

information to the public that they can use to estimate their personal risks (e.g., the risk associated with

consuming an 8-ounce meal of fish).

For each scenario, contaminant concentrations specific to an exposure medium that a receptor may

contact, called exposure point concentrations (EPCs), were calculated for each COPC; EPCs for COCs in

fish and shellfish are shown in Table 7.

Table 6. Rationale for Selection or Exclusion of Exposure Pathways in the Human Health Risk Assessment

Exposure

Scenario

Exposure

Point

Exposure

Medium

Receptor

Population

Receptor

Age

Exposure

Route

Type of

Analysis

Rationale for Selection or Exclusion of

Exposure Pathway

Water recreation

Water recreation areas in LDW

sediment resident

adult dermal,

ingestiona qualitative

Exposure via swimming less than exposure via beach play.

child dermal,


Exposure via swimming less than exposure via beach play.

surface water

resident

adult dermal,

ingestiona numeric Most likely direct contact pathway for surface water.

child dermal,

ingestiona numeric Most likely direct contact pathway for surface water.

Beach play in intertidal areab

LDW beaches

sediment resident

adult dermal,


Adult’s exposure during beach play likely to be less than child’s exposure on a per kilogram body weight basis.

child dermal,

ingestiona numeric

Residents may play at the shoreline near or adjacent to their houses.

surface water

resident

adult dermal,


Exposure attributable to resuspended sediment in water column is insignificant compared to that from direct contact with bedded sediment.

child dermal,


Exposure attributable to resuspended sediment in water column is insignificant compared to that from bedded sediment.


41

Exposure

Scenario

Exposure

Point

Exposure

Medium

Receptor

Population

Receptor

Age

Exposure

Route

Type of

Analysis

Rationale for Selection or Exclusion of

Exposure Pathway

Human consumption of resident seafood

Fishing/ shellfishing locations in the LDW

resident fish and shellfish tissue

resident, visitor, worker

adult, child

ingestion numeric

Although available data suggest current seafood consumption from LDW is low, tribal members have treaty harvest rights and the public also has recreational expectations for a fishable and swimmable estuary.

Fishing/ shellfishing in intertidal areas

Fishing locations in the LDW

sediment resident, visitor, worker

adult dermal,

ingestiona numeric

Recreational clamming may occur, given the abundance of clams in some areas. Incidental exposure during fishing is insignificant.

child dermal,


Incidental exposure during fishing likely to be less than that assumed in beach play scenario; potential exposure during clamming likely to be much lower compared to adult exposures.

surface water

resident, visitor, worker

adult dermal,

ingestiona qualitative Incidental exposure is insignificant.

child dermal,

ingestiona qualitative Incidental exposure is insignificant.

Occupational exposure (netfishing)

Commercial netfishing locations in LDW, which potentially include all LDW sediments

sediment

worker

adult dermal,

ingestiona numeric

Commercial fishers are active at the site throughout the fishing season; nets contact the sediment.

surface water

adult dermal,


Exposure attributable to resuspended sediment in water column is insignificant compared to that from bedded sediment.

Other occupational exposurec

industrial facilities adjacent to LDW

sediment worker adult dermal, ingestiona

qualitative Exposure expected to be much less than that evaluated in other sediment exposure scenarios.

surface water

worker adult dermal, ingestiona

qualitative Exposure expected to be much less than that evaluated in other scenarios.

a. Incidental ingestion associated with sediment contact. b. Although the beach play scenario is expected to be protective of adults who may participate in beach play activities, they may receive

exposure through other activities, such as dog walking. Thus, a dog-walking scenario is evaluated in the uncertainty analysis. c. Alternative occupational exposure scenarios were evaluated in the HHRA uncertainty section (LDWG 2010), including exposure scenarios

for a habitat biologist, habitat restoration volunteers, and King County special operations staff.


42

Table 7. Exposure Point Concentrations (EPCs) for Contaminants of Concern in Fish and Shellfish Used in Human Health Risk Assessment

Maximum Detected Concent.

Maximum Reporting

Limit

Number Detected/ Total Number of

Samples

EPC Value Statistica Rationaleb

Total PCBs, mg/kg ww

benthic fish, fillet 2.0 NA 33/33 1.2 c UCL95 95% Chebyshev, pooled RL

benthic fish, whole body

4.7 NA 45/45 2.6 UCL95 Approximate Gamma UCL

clams 0.58 NA 14/14 0.60 UCL95 99% Chebyshev (Mean, sd) UCL

crab, edible meat 0.39 0.020 26/29 0.20 UCL95 95% KM (t) UCL

crab, whole body 1.9 NA 25/25 1.1 UCL95 95% H-UCL

musselsd 0.060 0.013 18/22 0.041 UCL95 95% KM (Percentile Bootstrap) UCL

pelagic fish, whole body

18.4 NA 53/53 1.9 UCL95 95% H-UCL

PCB TEQ, mg/kg ww

benthic fish, fillet 1.41 × 10-5 NA 8/8 1.17 × 10-5 UCL95 Student’s-t UCL


2.47 × 10-5 NA 8/8 2.04 × 10-5 UCL95 Student’s-t UCL

clams 5.65 × 10-6 NA 8/8 3.16 × 10-6 UCL95 Approximate Gamma UCL

crab, edible meat 2.93 × 10-6 NA 8/8 2.41 × 10-6 UCL95 Student’s-t UCL

crab, whole body 1.16 × 10-5 NA 6/6 9.68 × 10-6 UCL95 Student’s-t UCL


7.30 × 10-5 NA 11/11 3.37 × 10-5 UCL95 Approximate Gamma UCL

Arsenic (Inorganic), mg/kg ww

benthic fish, fillet 0.006 0.003 6/8 0.0062 UCL95 95% Chebyshev, pooled ½ RL


0.09 NA 8/8 0.073 UCL95 Student’s-t UCL

clams 3.27 NA 8/8 2.0 UCL95 Student’s-t UCL

crab, edible meat 0.03 NA 6/6 0.042 UCL95 95% Chebyshev (Mean, sd) UCL

crab, whole body 0.123 NA 6/6 0.11 UCL95 Student’s-t UCL


0.16 0.01 8/10 0.088 UCL95 95% KM (t) UCL

benthic fish, fillet 0.0013 0.0079 1/11 0.004 one-half

maximum RL

Low number of detected samples

Carcinogenic PAH, mg/kg ww e

benthic fish, fillet 0.00064 0.00045 5/8 0.00064 maximum detect

Low number of samples


0.0028 0.00045 21/24 0.0023d UCL95 95% KM (Chebyshev) UCL


43

Maximum Detected Concent.

Maximum Reporting

Limit

Number Detected/ Total Number of

Samples

EPC Value Statistica Rationaleb

clams 0.044 NA 14/14 0.020 UCL95 approximate Gamma UCL

crab, edible meat 0.00084 0.00065 8/19 0.00065 UCL95 95% KM (t) UCL

crab, whole body 0.0024 NA 19/19 0.00092 UCL95 95% modified-t UCL


0.0022 NA 26/26 0.00095 UCL95 95% modified-t UCL

Dioxin TEQ, mg/kg wwf

NA NA NA NA NA NA

a. EPC statistics were calculated as follows: if there were no detected values, one-half of the maximum detection limit was used; if there were 1-5 detected values, the higher of one-half the maximum reporting limit or the maximum detected value were use; if 6 or more detections, ProUCL version 4.0 was used, with nondetected and detected values considered according to its algorithms.

b. Rationales for selecting appropriate statistics for EPC values are described in ProUCL version 5.0 (EPA 2013b) c. Because of the availability of historical data for English sole from RM 0-1.5, the EPC was calculated as a weighted mean, rather than a

mean of all data combined. Means were calculated for each of the four tissue sampling areas, with the historical data for English sole included in Area 1. The mean for the EPC derivation was then calculated as the arithmetic average of the four tissue sampling area means. The upper 95% confidence limit on that mean was estimated using Chebyshev's nonparametric method with a pooled standard deviation from the four areas. Thus, no assumption is made of equal means or variances across tissue sampling areas.

d. No mussel data were available for cPAHs. When calculating the chronic daily intake and risk values, seafood consumption that had been assigned to mussels was divided proportionally among the remaining consumption categories.

e. cPAH concentrations are given in terms of benzo(a)pyrene equivalents. Data used in the risk characterization section of this document are from only 2004 because of high reporting limits in historical data.

f. No dioxin tissue data were available for the HHRA, see discussion Section 5.2.1.

7.1.2.1 Consumption of resident seafood from the LDW

The HHRA considered consumption by groups that eat more fish and shellfish on average than the

general population. Table 8 displays the assumptions associated with exposure to contaminated seafood

for Tribal members (adults and children), and Asian Pacific Islander (API) adults used to estimate

consumption rates.12

The seafood consumption rates used in the HHRA were based on studies specific to

these two groups:

Tribal members — Puget Sound tribes have been shown to have some of the highest seafood

consumption rates of any regional population. To determine a resident-seafood consumption rate

for Tribal members, EPA analyzed (EPA 2007) a published survey of the Tulalip Tribes’

consumption of fish and shellfish from the Puget Sound region (Toy et al. 1996). Based on that

survey, an adult Tribal RME seafood consumption scenario equal to the 95th percentile of Puget

Sound seafood consumption was assumed equal to 97.5 g per day (approximately 13 8-ounce [oz]

meals per month). This seafood consumption rate does not include consumption of salmon.

Although salmon are a highly preferred and consumed fish from the LDW, human health risks

were not calculated for the consumption of salmon because most of the bioaccumulative

contaminants in salmon are acquired during their open ocean feeding, which is a long period

relative to their short LDW residence time. Additionally, an upper-bound (UB) Tribal risk

12

EPA estimates risks based on seafood consumption rates that would be in effect assuming that no contamination

was present. It would be inappropriate to use current seafood consumption rates from the LDW, because they

are likely suppressed due to general knowledge that the site is contaminated, and published and posted seafood

consumption advisories recommending no consumption of LDW resident fish or shellfish.


44

estimate was derived using estimated Tribal consumption rates based on Suquamish Tribe

seafood consumption data.13

Asian Pacific Islanders — Asian Pacific Islanders also have high seafood consumption rates

relative to the general population. For adult API, a resident seafood consumption rate of 51.5

g/day (approximately 7, 8-oz meals per month; also excluding salmon) was assumed based on a

1999 EPA study of API seafood consumption rates in King County (Sechena, et al. 1999).

Central tendency (CT) or average risks were also estimated for these groups.

Finally, the risk assessment provided an informational (I) scenario assuming consumption of one meal per

month of individual seafood categories. The informational scenario makes it easier to compare scenarios

when assessing the risks that would result from consuming one meal per month of different species found

in the LDW.

Table 8. Summary of Seafood Ingestion Exposure Parameters for Different Exposure Scenarios

Scenario a Total Ingestion Rate (g/day)b

Meals per Month c Exposure Duration (years)

Adult tribal (Tulalip data) - RME 97.5 13.1 70

Adult tribal (Tulalip data) – CT 15 2 30

Adult tribal (Suquamish data) - UB 584.2d 78 70

Child tribal (Tulalip data) - RME 39 13 6

Child tribal (Tulalip data) - CT 6.0c 2 6

Adult API – RME 51.5 6.9 30

Adult API – CT 5.3 0.7 12

Adult – Ie 7.5d 1 30 a. RME — reasonable maximum exposure. CT — central tendency. UB — upper bound. I — informational.

b. Rates include pelagic fish (shiner surfperch, striped and pile perch), benthic fish (English sole, starry flounder), crabs (Dungeness, slender and red rock), and clams and mussels. They do not include consumption of anadromous fish, which migrate through the LDW and do not reside long enough to be significantly affected by LDW contamination.

c. One meal is assumed to be equal to 227 g (8 oz) for adults and 90.8 g (3.2 oz) for children.

d. Totals of benthic and shellfish categories (and overall total) differ slightly from values provided by EPA (2007), because of significant figure and rounding issues when total consumption was allocated to more specific consumption categories.

e. Adult I (one-meal-per-month) scenario assumed 7.5 g/day for each of the individual seafood categories to reflect different fishing and consumption practices.

7.1.2.2 Direct contact with sediment

Risks were evaluated for exposure to contaminated sediment through both dermal (skin) contact and

incidental ingestion during commercial netfishing (adults), clamming (adult Tribal members and

recreational users), and beach play (children).

The risk associated with exposure to contaminants in surface water while swimming was not evaluated in

the same way as other exposure scenarios in the HHRA. Swimming risks were characterized using a 1999

King County analysis (King County 1999) for the LDW and Elliott Bay. Swimming-related excess cancer

risks were determined to be low, less than 1x10-6

.

13

The Suquamish Tribe fishes in the vicinity of the LDW and requested that Tribal risks used in the HHRA be based

on their consumption rates also, which they report are substantially higher than the rates reported in Toy et al

(1996). Suquamish Tribe consumption rates were considered part of a UB scenario. However, they were not

used as an RME scenario because Suquamish Tribe members consume a large amount of shellfish, and EPA has

determined that the LDW does not contain a large amount of high quality shellfish habitat.


45

Table 9 displays the exposure parameters used to evaluate exposures related to direct contact with

contaminated sediment, which included dermal (skin) contact and incidental ingestion of sediment.

Determination of areas associated with exposure was related to how receptors come into contact with

contaminants while engaging in a particular activity. It was assumed that netfishers would come into

contact with sediments throughout the LDW; for clamming, clam habitat areas accessible by boat or shore

were considered. Table 10 shows EPCs for beach play activities over discrete areas of intertidal sediment

accessible from the shoreline (outlined in blue in Figure 6). This approach was based on the assumption

that small children would contact sediment over limited areas and not all intertidal sediments. EPCs and

risks for beach play initially estimated in the HHRA were recalculated in the FS because additional data

became available for beach areas following completion of the HHRA. The updated FS dataset was not

used to recalculate risks associated with other pathways because the newer data did not significantly alter

them.

Table 9. Summary of Exposure Parameters for Direct Contact with Sediment for Different Exposure Scenarios

Scenarioa Exposure Frequency (days/year)

Incidental Sediment Ingestion Rate (g/day)

Exposure Duration (years)

Adult netfishing — RME 119 0.05 44

Adult netfishing — CT 63 0.05 29

Child (0-6 yrs) beach playb — RME 120 0.2 6

Tribal clamming (Adult) — RME 120 0.1 64

Tribal clamming (Adult) — UB 183 0.1 70

Adult clamming — I 7 0.1 30 a. RME — reasonable maximum exposure. CT — central tendency. UB — upper bound. I — informational.

b. For the beach play scenarios, the LDW was divided into eight areas to assess risks associated with different parts of the LDW.

Table 10. Summary of Sediment Exposure Point Concentrations (EPCs) for Contaminants of Concern Used in Human Health Risk Assessment and Updated in the Feasibility Study

Exposure Pointa,c

Concentrations Detected (sediment) EPC

Maximum Detected Concent., mg/kg dw

Maximum Reporting

Limit, mg/kg dw

Number Detected/Total

Number of Samples

EPC Value,

mg/kg dw

Statistic Usedb

Total PCB

beach play RME, area 1 0.09 NA 2/2 composites 0.05 weighted composite samples

beach play RME, area 2 0.56 0.02 7/8 0.29 95% KM (t) UCL

beach play RME, area 3 0.42 0.02 14/18 0.22 95% KM (Chebyshev) UCL

beach play RME, area 4 2900 0.04 28/29 1100 99% KM (Chebyshev) UCL


Duwamish Waterway Park (subset of beach play area 5) 0.28 NA 1/1 composite 0.28 composite sample

beach play RME, area 6 0.86 NA 1/1 composite 0.86 composite sample

beach play RME, area 7 0.34 0.04 16/22 0.09 95% KM (BCA) UCL

beach play RME, area 8 0.52 0.04 15/22 0.10 95% KM (BCA) UCL

tribal clamming RME/ UB, accessible by bank or boat 110 0.040 415/440 4.00 97.5% KM (Chebyshev) UCL

recreational clamming I, on bankside accessible beaches 23 0.040 142/161 1.50 97.5% KM (Chebyshev) UCL

Netfishing RME/CT, waterway wide 220 0.05 1,205/1,291 2.50 97.5% KM (Chebyshev) UCL


46

Exposure Pointa,c



Maximum Reporting

Limit, mg/kg dw


Number of Samples

EPC Value,

mg/kg dw

Statistic Usedb

PCB TEQ

beach play RME, area 1 9.08 × 10-8 NA 1/1 9.08 × 10-8 maximum detect


beach play RME, area 3 NA NA na na na






tribal clamming RME/UB, accessible by bank or boat 1.38 × 10-3 NA 30/30 1.84 x 10-5 97.5% Chebyshev (MVUE) UCL

recreational clamming I, on bankside accessible beaches 2.04 ×10-4 NA 18/18 4.19 × 10-5 99% Chebyshev (MVUE) UCL

Netfishing RME/CT, waterway wide 1.38 × 10-3 NA 48/48 7.18 x 10-5 95% Chebyshev (MVUE) UCL

Arsenic (Inorganic) beach play RME, area 1 25.3 NA 2/2 composites 16 Weighted composite samples

beach play RME, area 2 20.7 NA 6/6 19 95% Student's t UCL

beach play RME, area 3 18.3 6.6 10/13 11 95% KM (Percentile Bootstrap) UCL

beach play RME, area 4 48.7 NA 25/25 12 95% approximate gamma UCL

beach play RME, area 5 19.1 NA 26/26 10 95% approximate gamma UCL

Duwamish Waterway Park (subset of beach play area 5) 4.3 4.3 1/1 composite 4.3 composite sample

beach play RME, area 6 93.8 NA 1/1 composite 94 composite sample

beach play RME, area 7 13.8 NA 14/14 10 95% Student's t UCL

beach play RME, area 8 15.6 NA 1.0 9 95% approximate gamma UCL

tribal clamming RME/UB, beaches accessible by bank or boat 1,100 31.0 254/275 27 95% KM (BCA) UCL

recreational clamming I, on bankside accessible beaches 20.7 8.8 100/103 10 95% KM (Percentile Bootstrap) UCL

Netfishing RME/CT, waterway wide 1,100 31 755/817 21 95% KM (BCA) UCL

cPAH TEQ

beach play RME, area 1 0.38 360 2/2 composites 0.38 weighted composite samples

beach play RME, area 2 3.0 NA 6/6 7.0 95% KM (Chebyshev) UCL


beach play RME, area 4 4.80 0.018 23/25 1.4 97.5% KM (Chebyshev) UCL

beach play RME, area 5 0.001 NA 1.0 0.4 95% Chebyshev (MVUE) UCL

Duwamish Waterway Park (subset of beach play area 5) 0.061 NA 1/1 composite 0.27 composite sample

beach play RME, area 6 7.1 NA 1/1 composite 7.1 composite sample

beach play RME, area 7 0.073 0.017 12/14 0.098 95% KM (t) UCL

beach play RME, area 8 0.62 NA 14/15 0.32 95% Students-t UCL

tribal clamming RME/UB, beaches accessible by bank or boat 11.0 0.11 255/264 0.77 95% KM (Chebyshev) UCL

recreational clamming I, on bankside accessible beaches 3.0 0.036 97/103 0.48 95% KM (Chebyshev) UCL

Netfishing RME/CT, waterway wide 11.0 0.1 749/793 0.6 95% KM (Chebyshev) UCL


47

Exposure Pointa,c



Maximum Reporting

Limit, mg/kg dw


Number of Samples

EPC Value,

mg/kg dw

Statistic Usedb

Dioxin/Furan TEQ

beach play RME, area 1 2.77 × 10-6 NA 2/2 composites 2.5 × 10-6 weighted composite samples




beach play RME, area 5 3.57 × 10-5 NA 4/4 (2 composites

and 2 grab samples) 3.57 × 10-5 maximum detect

Duwamish Waterway Park (subset of beach play area 5) 6.28 × 10-6 NA 1/1 composite 6.28 × 10-6 composite sample

beach play RME, area 7 3.73 × 10-6 NA 1/1 composite 3.73 × 10-6 composite sample

beach play RME, area 8 3.79 × 10-6 NA 1/1 3.79 × 10-6 EPC based on a single grab sample.

tribal clamming RME/UB, accessible by bank or boat 2.1 × 10-3 NA 11/11 1.42 × 10-3 95% Adjusted Gamma UCL

Recreational clamming I, on bankside accessible beaches 4.1 × 10-4 NA 6/6 3.65 × 10-4 95% Chebyshev (MVUE) UCL

Netfishing RME/CT, waterway wide 2.1 × 10-3 NA 43/43 6.1 x 10-4 99% Chebyshev (Mean, sd) UCL

a. Beach play values are based on FS dataset; clamming and netfishing are based on RI dataset b. The recommended statistical methods used for calculating the EPC for the dioxin contaminants were based on recommendations from

ProUCL (most recent version is EPA 2013b) c. RME represents an exposure frequency of 120 days/year; CT represents an exposure frequency of 63 days/year; UB represents 183

days/year; and I represents 7 days/year.

7.1.3 Toxicity Assessment

The toxicity assessment evaluated the relationship between the magnitude of exposure to a contaminant

(i.e., the dose) and the risk or hazard posed to humans. Toxicity assessment utilizes human or animal data

to predict human toxicity. Scientific estimates of toxicity are adjusted to account for uncertainty so that

they are health protective. The toxicity assessment contains two steps: hazard characterization, and dose-

response evaluation. The assessment provided, when possible, a numerical estimate of the increased

likelihood of adverse effects associated with exposure to a specific dose of a contaminant.

Hazard characterization identifies the types of toxic effects a contaminant can exert. “Toxicity values” are

the numerical expressions of predicted toxic effects. Health risks are calculated differently for

carcinogenic and noncarcinogenic effects, and separate toxicity values have been developed accordingly.

The primary source of toxicity values (cancer slope factors [SFs] and noncancer reference doses [RfDs])

is EPA’s Integrated Risk Information System (IRIS) database. If a toxicity value was not available from

IRIS for a given contaminant, then the toxicity value used was from EPA’s Office of Research and

Development/National Center for Environmental Assessment, they are called Provisional Peer-Reviewed

Toxicity Values. If neither of these sources provided a toxicity value, additional EPA and non-EPA

sources of toxicity information were used, including information from EPA regional offices, EPA Health

Effects Assessment Summary Tables (HEAST) values, information from California EPA, and Agency for

Toxic Substance and Disease Registry (ATSDR) Minimal Risk Levels. Table 11 and Table 12 display the

cancer and noncancer toxicity data used, respectively.


48

The toxicity of cPAHs, dioxins/furans and dioxin-like PCB congeners is expressed as a toxic equivalency

factor or TEQ. (Congeners are unique individual chemicals in a group such as PCBs or dioxins/furans.)

The TEQ expresses the toxicity of each contaminant in a structurally similar group relative to a reference

compound – 2,3,7,8-tetrachloro-p-dibenzodioxin (2,3,7,8-TCDD) for dioxins/furans and benzo(a)pyrene

for cPAHs.

Children may be more sensitive to the effects of certain contaminants than adults. For example, EPA

accounts for the enhanced toxicity of cPAHs to children by using age-dependent adjustment factors.

The dose-response evaluation examines the level of toxic effect associated with a specific amount of a

contaminant. The magnitude of a toxic response to a contaminant depends on the dose to a receptor. The

time period of exposure is an important consideration in toxicity studies. For most Superfund sites, EPA

is concerned with long term or chronic exposure to contaminants.

7.1.4 Risk Characterization

Results of the exposure assessment (estimated contaminant intakes or doses) are combined with the

results of the dose-response assessment (toxicity values established in the toxicity assessment) to provide

numerical estimates of potential health effects. The quantification approach differs for potential cancer

and noncancer effects, as described below.

Cancer

The potential for cancer effects is evaluated by estimating the excess lifetime cancer risk (ELCR). This

risk is the incremental increase in the probability of developing cancer during one’s lifetime in addition to

the background probability of developing cancer. Cancer slope factors developed by EPA are considered

to be a plausible upper-bound estimate of the cancer potency of a contaminant. By using these upper-

bound estimates for the cancer slope factors, there is reasonable confidence that the actual cancer risk will

not exceed the estimated risks and may actually be lower (EPA 1989).

For the LDW Site, ELCRs were estimated using the following equation:

Risk = CDI × SF

Where:

Risk = Excess lifetime cancer risk (unitless probability)

CDI = Chronic daily intake averaged over a lifetime (mg/kg-day)

SF = Cancer slope factor (mg/kg-day)-1

When multiple contaminants are present, cancer risks are added together for each exposure scenario (e.g.,

fish consumption and clamming). Risks may be summed across exposure scenarios (e.g. seafood

consumption risk and clamming direct contact risks). This is consistent with the EPA risk assessment

guidelines (EPA 1989).

Table 11 shows cancer toxicity values for the COCs. Accessed dates reflect information available at the

time of the risk assessment.


49

Table 11. Cancer Toxicity Data Summary for Human Health COCs

Contaminant of Concern

Oral/Dermal Cancer Slope Factor, (mg/kg-d)-1

Weight of Evidence/Cancer Guidelines Descriptiona

Source Information Accessed Date

Total PCB 2 B2 IRIS (upper-bound slope factor specified for use with persistent and bioaccumulative PCBs)

3/8/2006

PCB TEQ 150,000 B2 HEAST (slope factor based on 2,3,7,8-TCDD) 3/8/2006

Arsenic 1.5 A IRIS (surrogate for inorganic arsenic) 3/8/2006

cPAH 7.3 B2 IRIS (benzo[a]pyrene toxicity equivalents) 3/8/2006

Dioxin/Furan 150,000 A (IARC) HEAST 4/7/2006

a. A = known human carcinogen; B2 = probable human carcinogen (based on sufficient evidence in animals and inadequate or no evidence in humans). The full set of chemicals in each of the classes represented is available at http://www.greenfacts.org/glossary/def/epa-cancer-classification.htm.

Noncancer

For noncancer effects, the likelihood that a receptor will develop an adverse effect is estimated by

comparing the predicted level of exposure for a particular contaminant with the highest level of exposure

that is considered protective (i.e., its RfD). When the ratio exceeds 1 (i.e., exposure exceeds RfD), there is

a concern for potential noncancer health effects. The ratio of the chronic daily intake (CDI) divided by

RfD is termed the hazard quotient (HQ):

HQ = CDI / RfD

To screen for potential for noncancer effects associated with exposure to multiple contaminants, a hazard

index (HI) is used (EPA 1989). This summation approach assumes that the noncancer, multiple-

contaminant hazard is additive, which may not always be the case (as when dissimilar organ systems are

affected).

The HI is calculated as follows:

HI = ∑ 1 Ei /RfDi

Where:

HI = hazard index

Ei = daily intake of the ith contaminant (mg/kg-day)

RfDi = reference dose of the ith contaminant (mg/kg-day)

N = number of contaminants

If the overall HI exceeds 1, EPA will evaluate HIs for specific organ systems or toxic effects to determine

if exposure to multiple contaminants is truly of concern.

Table 12 shows the noncancer toxicity values for the COCs.

N

http://www.greenfacts.org/glossary/def/epa-cancer-classification.htm

http://www.greenfacts.org/glossary/def/epa-cancer-classification.htm


50

Table 12. Noncancer Toxicity Data Summary for Human Health

Contami-nants of Concern

Chronica or Subchronicb

Exposure

Oral RfD, mg/kg-day

Value

Oral Absorption Efficiency for

Dermal Exposure Primary Target Organ

Combined Uncertainty

and Modifying Factors

Sources

Information Accessed

Date Value Reference

Arsenic Chronic 0.0003 0.03 EPA 2004 Cardiovascular system; Skin (hyperpigmen-tation, keratosis)

3 IRIS (surrogate for inorganic arsenic)

3/8/2006

Total PCB Chronic 0.00002 0.14 EPA 2004 Development, Immune system, nervous system

300 IRIS (surrogate = Aroclor 1254, the lowest and most protective RfD available for PCBs in IRIS; total includes Aroclors 1016, 1221, 1232, 1242, 1248, 1254, and 1260)

3/8/2006

Source: EPA. 2004. Risk assessment guidance for Superfund, Volume 1: Human health evaluation manual (Part E, supplemental guidance for dermal risk assessment). Final, July 2004. EPA/540/R/99/005. Office of Emergency and Remedial Response, US Environmental Protection Agency, Washington, DC.

IRIS = Integrated Risk Information System a. Chronic represents exposure of a population (including sensitive individuals) that occurs on a daily basis, which is likely to be without

appreciable risk of deleterious effects during a lifetime (EPA 1989). b. Subchronic represents exposure of a population (including sensitive individuals) between 2 weeks and 7 years, and is likely to be without

appreciable risk of deleterious effects (EPA 1989).

Risk Characterization and Selection of Human Health COCs

Table 13 shows estimated cancer and noncancer risks for the human health scenarios. Figure 8 and Figure

9 summarize the baseline excess cancer and noncancer risk related to the number and type of seafood

meals consumed per month. Figure 10 summarizes the excess cancer and noncancer risk associated with

the RME scenarios for direct sediment contact scenarios. Site seafood consumption RME risks exceed

direct contact risks. Seafood consumption risks and direct contact HQs for the RME scenarios exceed risk

thresholds established under CERCLA. Under CERCLA, these thresholds are excess cancer risks of

1 x 10-4

and a noncancer HQ (or HI) of 1. MTCA thresholds are also exceeded, which are excess cancer

risks of 1 x 10-6

for individual contaminants or 1 x 10-5

cumulatively for multiple contaminants, and a

noncancer HQ or HI of 1. For direct contact scenarios, HQs were less than 1, with the exception of beach

play at Beach Area 4.

Dioxins and furans were not included in the total excess cancer risk calculation for the RME seafood

consumption scenarios because data on dioxin/furan concentrations in fish and shellfish tissue were not

collected during the RI. These data were not gathered during the RI because data already available

indicated that dioxin/furan concentrations in most Puget Sound fish and shellfish would present

unacceptable risk at the RME consumption rates; therefore, additional data were not needed and the

HHRA assumed unacceptable risks due to dioxins/furans in the LDW without further investigation.

However, in May 2007, after the HHRA was finalized, Ecology sampled and analyzed a few skin-off

English sole fillets collected near Kellogg Island. Data from these samples were used to calculate the

excess cancer risks associated with dioxins/furans as 6 x 10-5

for the adult Tribal RME scenario. This risk

estimate is uncertain because it is based on a smaller number of samples than in datasets typically used for

an HHRA and is from a very limited portion of the LDW. It also does not include dioxin concentrations

for all seafood species used in the HHRA. Nevertheless, it provides some information regarding

dioxin/furan risks relative to risks from other COCs.

Table 14 summarizes the rationale for selection of human health COCs. Although BEHP,

pentachlorophenol, vanadium, tributyltin, and several pesticides were found in the waterway at


51

concentrations that exceeded risk thresholds, they were not selected as COCs due to low detection

frequency, low contribution to overall risk, or quality assurance concerns with analytical data.

Information on whether a contaminant was historically used at the Site was also considered in

determining whether these contaminants should be selected as COCs. PCBs, arsenic, cPAHs, and

dioxins/furans were identified as human health COCs based on an excess cancer risk greater than 1 x 10-6

for carcinogens, or a hazard quotient (HQ) greater than 1 for noncarcinogens. Other COPCs that exceeded

risk thresholds but were not designated as COCs were still evaluated in the FS to ensure that a cleanup

based on the COCs would also address risk due to these other contaminants.

Table 13. Summary of Cancer and Noncancer Risk Estimates for Human Health Scenarios

Scenarioa Medium Contaminant of Concern Excess Cancer Risk Hazard Quotient

Seafood Consumption Scenarios

Adult Tribal Seafood RME Consumption - Tulalip Survey

Fish and Shellfish

PCBs 2 x10-3 40

Inorganic arsenic 2 x10-3 4

cPAHs 8 x10-5 nc

Otherc 4 x 10-4 2

Total 4 x 10-3 nc

Adult Tribal Seafood CT Consumption - Tulalip Survey

Fish and Shellfish

PCBs 6 x10-5 4 Inorganic arsenic 7 x10-5 0.4 cPAHs 4 x10-6 nc Other 1 x10-5 0.2 Total 1 x10-4 nc

Adult Tribal Seafood UB Consumption - Suquamish Survey

Fish and Shellfish

PCBs 1 x10-2 274 Inorganic arsenic 2 x10-2 38 cPAHs 8 x10-4 nc Otherc 3 x10-3 15 Total 3 x10-2 nc

Child Tribal Seafood RME Consumption - Tulalip Survey

Fish and Shellfish

PCBs 3 x 10-4 87

Inorganic arsenic 3 x10-4 8

cPAHs 8 x 10-5 nc

Otherc 8 x 10-5 3

Total 8 x 10-4 nc

Child Tribal Seafood CT Consumption - Tulalip Survey

Fish and Shellfish


Adult Asian Pacific Islander RME Seafood Consumption

Fish and Shellfish

PCBs 5 x 10-4 29

Inorganic arsenic 7 x 10-4 3

cPAHs 3 x 10-5 nc

Otherc 1 x 10-4 1

Total 1 x 10-3 nc

Adult Asian Pacific Islander CT Seafood Consumption

Fish and Shellfish


Adult Informational (one meal per month)d

pelagic fish PCBs 2 x 10-4 10 Clam Inorganic arsenic 1 x 10-4 0.7 Clam cPAHs 7 x10-6 nc benthic fish (cancer)/ clam (HQ) Other 2 x10-5 0.3 clam/pelagic fish Total 2 x 10-4 nc


52

Scenarioa Medium Contaminant of Concern Excess Cancer Risk Hazard Quotient

Direct Contact Scenarios

Netfishing RME (Direct Sediment Contact)

Subtidal and Intertidal Sediment

PCBs 2 x 10-6 < 1

Arsenic 6 x 10-6 < 1

cPAHs 1 x 10-6 nc

Dioxins/Furans 2 x 10-5 nc

Other 2 x 10-6 < 1

Total 3 x 10-5 nc

Netfishing CT (Direct Sediment Contact)

Subtidal and Intertidal Sediment

PCBs 3 x 10-7 < 1 Inorganic arsenic 1 x 10-6 < 1 cPAHs 2 x 10-7 nc Dioxins/Furans 4 x 10-6 nc Other 3 x 10-7 < 1 Total 5 x 10-6 nc

Clamming RME (Direct Sediment Contact)

Intertidal Sediment

PCBs 8 x 10-6 < 1

Arsenic 2 x 10-5 < 1

cPAHs 5 x 10-6 nc

Dioxins/Furans 1 x 10-4 nc

Other 6 x 10-6 < 1

Total 1 x 10-4 nc

Clamming UB (Direct Sediment Contact)

Intertidal Sediment

PCBs 1 x 10-5 < 1 Inorganic arsenic 3 x 10-5 < 1 cPAHs 8 x 10-6 nc Dioxins/Furans 2 x 10-4 nc Other 9 x 10-6 nc Total 3 x 10-4 < 1

Clamming I (Direct Sediment Contact)

Intertidal Sediment

PCBs 9 x 10-8 < 1 Inorganic arsenic 3 x 10-7 < 1 cPAHs 1 x 10-7 nc Dioxins/Furans

8 x 10-7 nc Other 2 x 10-8 nc Total 1 x 10-6 < 1

Beach Play RME (Direct Sediment Contact -Ranges for 8 beaches)

Intertidal Sediment

PCBs 3 x 10-8 to 6 x 10-4

< 1b

Arsenic 3 x 10-6 to 3 x 10-5

< 1

cPAHs 1 x 10-6 to 8 x 10-5

nc

Dioxins/Furans 1x 10-7 to 1x 10-5

nc

Total 4 x 10-6 to 6 x 10-4

nc

Notes: General

nc = not calculated

Baseline lifetime excess cancer risks: calculated as the sum of the risk estimate for inorganic arsenic, cPAHs, and PCBs. Estimates for seafood consumption scenarios do not include risk estimates from dioxins and furans.

The ”Other”’ category includes those contaminants evaluated in the HHRA with concentrations greater than screening levels (i.e., both those that exceed risk thresholds and some that do not).

a. For the netfishing and clamming RME scenarios, the total excess cancer risks are based upon data from the Remedial Investigation only. For the beach play RME scenarios, the estimation of total excess cancer risks are based upon data from the Remedial Investigation along with additional data collected during the Feasibility Study.

b. All beaches but Beach Play Area 4 at RM 2.2 have HQs <1. The Beach Play Area 4 HQ of 2 excludes two very high PCB concentrations (see footnote a to Table 1); if the two high PCB concentrations were included, the HQ would be 187. Beach Play Areas are shown in Figure 6.

c. HQ > 1 was due to tributyltin (TBT). d. Informational exposure is presented for the food groups with the highest associated risks.


53

Notes:

Excess cancer risks and noncancer hazard quotients (HQs) were calculated using the exposure assumptions for the adult tribal RME seafood consumption scenario.

One meal is equal to 8 ounces, and 3 meals per week is approximately equal to the rate used for the adult tribal RME scenario in the HHRA.

Excess cancer risks were calculated as the sum of excess cancer risks for inorganic arsenic, cPAHs, and PCBs. These estimates do not include risk estimates from dioxins/furans, as discussed in Section 7.1.4. For calculating market basket consumption, the risks for cPAHs and inorganic arsenic are based only on the consumption of clams because clams account for over 95% of the risk.

Noncancer HQs for total PCBs are presented because HQs were the highest for PCBs.

Figure 8. Baseline Excess Cancer Risk and Noncancer Hazard Quotients for Consumption of Various Seafood Species as a Function of the Number of Meals Consumed per Month


54

Notes: One meal is 8 ounces for adults and 3.2 ounces for children. Baseline tissue dataset: baseline excess cancer risks and hazard quotients (HQs) are based on the tissue dataset in the RI, and were calculated as the sum of the risk estimates for arsenic, cPAHs,

and PCBs, and other contaminants. These estimates do not include risks from dioxins and furans. The “other” category includes those contaminants evaluated in the HHRA with concentrations greater than conservative screening levels. For baseline, only HQs for total PCBs are presented because non-cancer HQs were the highest for PCBs.

Figure 9. Baseline Noncancer Hazard Quotients and Excess Cancer Risk for the Seafood Consumption RME Scenarios


55

Notes: Baseline datasets: Total excess cancer risks are based upon the RI (LDWG 2010) dataset for the netfishing and clamming RME scenarios, and are based upon the FS (LDWG 2012a) dataset for the

beach play RME scenarios. Non-cancer HQs: HQs are not shown in this figure because they were less than 1 in all cases, except for the beach play area located at the head of the inlet at RM 2.2 west. If these 2 samples are

considered, the PCB HQ would be 187; when they are removed, the resulting HQ would be 2.

Figure 10. Baseline Excess Cancer Risk for the Direct Sediment Contact RME Scenarios


56

Table 14. Summary of COPCs and Rationale for Selection as COCs for Human Health Exposure Scenarios

COPC COC?

Maximum RME Risk Estimate Rationale for Selection or Exclusion as COC

Seafood Consumption Scenarios

PCBs Yes 2 × 10-3 Risk magnitude, high percent contribution to the cumulative excess cancer risk (58%), and high detection frequency in tissue samples (97%).

Inorganic arsenic Yes 2 × 10-3 Risk magnitude, percent contribution to the cumulative excess cancer risk (29%), and high detection frequency in tissue samples (100%).

cPAHs Yes 8 × 10-5 Risk magnitude and high detection frequency in tissue samples (72%).

Dioxins/furans Yes nd No dioxin/furan tissue data were available. However, because excess cancer risks were assumed to be unacceptably high, dioxins/furans were identified as a COC.

Bis(2-ethylhexyl) phthalate

No 6 × 10-6 Low percent contribution to the cumulative excess cancer risk (less than or equal to 3%) and rarely detected in tissue samples (particularly when samples were re-analyzed to evaluate the effect on RLs of analytical dilutions in the initial analysis). Pentachlorophenol No 9 × 10-5

Tributyltin No HQ = 3 HQs for these metals were only slightly greater than 1 and were driven by the child tribal RME seafood consumption scenario, for which Ingestion rates are uncertain. Vanadium No HQ = 2

Aldrin No 5 × 10-5

All organochlorine pesticides were low contributors to the cumulative excess cancer risk (less than or equal to 3% of the cumulative risk). In addition, because of analytical interference of these contaminants with PCBs, much of the tissue data for these contaminants were qualified JN, which indicates the presence of an analyte that has been ‘tentatively identified,’ and the associated numerical value represents its approximate concentration. The JN-qualified organochlorine pesticide results are highly uncertain and likely biased high.

alpha-BHC No 2 × 10-5

beta-BHC No 6 × 10-6

Carbazole No 4 × 10-5

Total Chlordane No 6 × 10-6

Total DDTs No 2 × 10-5

Dieldrin No 1 × 10-4

gamma-BHC No 5 × 10-6

Heptachlor No 1 × 10-5

Heptachlor epoxide No 3 × 10-5

Hexachlorobenzene No 1 × 10-5

Direct Sediment Exposure Scenarios

PCBs Yes 8 × 10-6 Lower risk magnitude and percent contribution to cumulative excess cancer risk than the other sediment risk drivers, but selected because of importance in the seafood consumption scenarios.

Inorganic arsenic Yes 2 × 10-5 Risk magnitude, percent contribution to cumulative excess cancer risk (14 to 19%), and high detection frequency in surface sediment samples (92%).

cPAHs Yes 4 × 10-5 Risk magnitude, percent contribution to cumulative excess cancer risk (3 to 85%), and high detection frequency in surface sediment samples (94%).

Dioxins/furans Yes 1 × 10-4 Risk magnitude, percent contribution to cumulative excess cancer risk (35 to 72%), and high detection frequency in surface sediment samples (100%).

Toxaphene No 6 × 10-6 Low percent contribution to cumulative excess cancer risk (6% or less) and low detection frequency in surface sediment samples (1%).

Notes: BHC = benzene hexachloride.

Except for TBT and Vanadium, the maximum RME risk estimates shown are excess cancer risks for the adult Tribal RME seafood consumption based upon Tulalip tribal data. Only RME scenarios were used to designate COCs. The highest risk estimate for any of the RME scenarios is shown in this table (adult tribal RME based on Tulalip data for seafood consumption, and various scenarios for direct contact). Note that the estimates reported here differ slightly from those reported in Appendix B (the HHRA) and Section 6 of the RI (LDWG 2010), based on a 2009 erratum (LDWG 2009) that adjusted the proportion of crabs and clams consumed by the Tulalip Tribe.


57

The majority of risks for seafood consumption were from PCBs and inorganic arsenic in resident fish,

crabs, and clams. While risks from PCBs were associated with all types of fish and shellfish evaluated,

the vast majority of risks due to inorganic arsenic and cPAHs (96-98%) were attributable to consumption

of clams. Lower risks were associated with activities that involve direct contact with sediment, such as

clamming, beach play, and netfishing. The majority of risk associated with adult direct sediment exposure

was from dioxins and furans. In contrast, the majority of carcinogenic risks for young children through

direct sediment exposure pathways (i.e., beach play) were from cPAHs, because cPAHs are more toxic to

young children than to adults. PCBs, arsenic, cPAHs, and dioxins/furans, along with the COCs identified

by the Ecological Risk Assessment (see Section 7.2), were selected as COCs and used to identify areas

requiring cleanup in the FS.

7.1.5 Uncertainty Analysis for the HHRA

To be health-protective of all members of the general public, the risk estimates presented in the baseline

HHRA were intended to avoid underestimation of risks for individuals with reasonable maximum

exposure (RME), and thus are likely to overestimate risks for most individuals for the contaminants that

were evaluated. EPA believes this is appropriate, as use of central tendency risk estimates in developing

cleanup actions would result in potentially unacceptable exposures following remediation. Central

tendency risk estimates were intended to provide information to risk managers involved in remedial

planning for the Site but may not reflect actual risks to people currently consuming LDW seafood.

Although risk estimates were highest for the seafood consumption scenarios, the uncertainties associated

with those risk estimates were also high. There is considerable uncertainty about the applicability of some

of the seafood consumption rates to this HHRA under current uses of the Site, particularly for clams,

given the degraded quality and quantity of shellfish habitat in the LDW. EPA’s risk analyses must also

account for potential future exposure. The Tribes involved in the LDW cleanup have indicated that they

would catch and consume more seafood from the LDW if contaminant concentrations were reduced.

Another uncertainty is in the methods used to characterize the excess cancer risks associated with

exposures to PCBs. Two methods were used in the baseline HHRA, one method based on total PCB data

and the cancer SF for total PCBs and a second method based on dioxin-like PCB congener data and the

cancer SF for 2,3,7,8 TCDD and toxic equivalency factors that describe how toxic each individual dioxin-

like PCB is relative to TCDD. The issue of concern is how to consider the joint risk posed by total PCBs

and dioxin-like PCBs. Some of the risk posed by dioxin-like PCBs is accounted for in the risk calculated

using total PCB dose and the PCB slope factor. Consequently, adding the dioxin-like PCB and total PCB

risks together could double-count the risk posed by dioxin-like PCBs and overestimate the risk posed by

PCBs. On the other hand, the bioaccumulation process may enrich levels of dioxin-like PCBs in seafood

relative to the levels of dioxin-like PCBs found in the commercial mixtures used to establish PCB

toxicity. Failure to account for this enrichment could underestimate the risk posed by PCBs. The true

risk posed by PCBs thus lies between the individual total PCB and dioxin-like PCB risk estimates and the

sum of these individual risk estimates. In the HHRA, the total PCB and dioxin-like PCB risk estimates

are presented separately along with the points noted above. Risk estimates in Table 13 are based on total

PCBs, as the risk accounted was generally greater using that method.

Dioxins and furans were not analyzed in seafood samples, as discussed in Section 7.1.4. This data gap

contributes to an underestimation of risk because these contaminants were not included in the risk

assessment for the seafood consumption scenarios.


58

The final risk estimates also reflect uncertainties associated with using data and assumptions from

multiple sources, including non-Site related sources such as background contributions of hazardous

substances; the combined effect of those uncertainties on risk estimates cannot be quantified. However,

the risk assessment tended to overestimate risks more than underestimate them, consistent with the health-

protective nature of risk assessment. In spite of these uncertainties, the baseline risk characterization for

the LDW site is considered to be protective of human health and sufficient to support risk management

decisions.

7.2 Ecological Risks

The baseline Ecological Risk Assessment (ERA) estimated risks for the benthic invertebrate, fish, crabs,

and wildlife species that may be exposed to contaminants in sediment, water, and aquatic biota in the

LDW. This assessment was based on historical data and sediment and tissue chemistry data collected as

part of the RI, as discussed in Section 5.3. The baseline ERA is an estimate of the likelihood of ecological

risks if no cleanup action is taken.

7.2.1 Ecological Communities in the LDW

Though much habitat has been lost, the LDW is home to a diverse ecology, with abundant resident and

non-resident fish and shellfish, bottom-dwelling organisms, marine mammals, and birds. As discussed in

Section 6.2, several LDW habitat restoration projects are planned over the next few years, which, along

with reduction of contamination, should result in increased waterway use by these organisms in the

future.

Several benthic fish (bottomfish such as sole, sculpin, and flounder) and pelagic fish (water column fish

such as perch and herring) are abundant in the LDW, as are salmon. The Green/Duwamish River system

supports eight species of salmonids: coho, Chinook, chum, sockeye, and pink salmon, plus cutthroat trout,

both winter- and summer-run steelhead, and bull trout. Juvenile Chinook and chum have a residence time

in the LDW from several days to two months; juvenile coho are in the LDW for only a few days; and

sockeye are rare in the LDW. Salmon found in the LDW spawn mainly in the middle reaches of the Green

River and its tributaries. The juvenile outmigration generally starts between March and June.

Outmigration usually lasts through mid-July to early August.

Puget Sound Chinook salmon are listed as threatened under the federal Endangered Species Act (ESA).

Other relevant fish species listed as threatened under the ESA include the coastal Puget Sound bull trout

and the Puget Sound steelhead. The LDW is designated as critical habitat for bull trout and Chinook

salmon. The bald eagle was delisted in 2007 under the ESA but is protected under the Bald and Golden

Eagle Protection Act, and under the Migratory Bird Treaty Act.

Typical of most estuaries, the benthic invertebrate community is dominated by annelids (worms),

mollusks (clams and snails), and crustaceans (e.g., shrimp and crabs). Dungeness and other crabs are

present in the LDW, although their distribution is generally limited to the portions of the LDW with

higher salinity.

The common shorebirds and wading birds observed in the LDW are sandpipers, killdeer, and great blue

herons. Bald eagles, ospreys, and great blue herons nest on or near the LDW and use the LDW for

foraging. The LDW provides habitat for mammal species including harbor seals, sea lions, and river

otters.


59

7.2.2 Problem Formulation

The problem formulation of the ERA established the overall scope of the assessment. It included a

description of the data available for conducting the ERA, the suitability of the data for risk assessment

purposes, and a risk-based screening process that allowed the risk assessment to focus on contaminants of

potential concern (COPCs) and eliminate contaminants that posed minimal risks to ecological receptors

from further consideration.

The ERA evaluated risks to four types of ecological receptors of concern (ROC) exposed to the

contaminants in the LDW, either directly or via ingestion of prey:

benthic invertebrates and crabs;

fish (juvenile Chinook salmon, Pacific staghorn sculpin, English sole);

birds (spotted sandpiper, great blue heron, osprey); and

wildlife (river otter and harbor seal).

These ROCs were selected to represent organisms with a range of characteristics that affect exposure,

such as habitat, dietary preferences, level in the food chain, and sensitivity to contaminants. Generally, if

these species are protected by the remedy, the many species they represent are also protected. Juvenile

salmon were selected because they are listed as a threatened species under ESA.

The problem formulation also presented conceptual site models for the ROCs (Figure 11). Conceptual site

models identify and describe pathways in which ROCs may be exposed to COPCs within the LDW. The

pathways evaluated in the ERA included both direct exposure through sediment and water and indirect

exposure through the ingestion of prey from the LDW. The potential exposure pathways of contaminants

to higher-trophic-level ROCs in the LDW is discussed in the RI and briefly summarized in Section 5.2.

7.2.3 Identification of Contaminants of Potential Concern for Ecological Receptors For each receptor of concern (ROC), COPCs were identified through a conservative, risk-based screening

process comparing maximum exposure concentrations to the numerical SCO for protection of benthic

invertebrates promulgated under the Washington Sediment Management Standards (benthic SCO), (Table

15) or, for other receptors, to no observed adverse effect levels (NOAELs) from the scientific literature.

COPCs identified included: 46 contaminants for the benthic invertebrate community (including

tributyltin [TBT], metals, and PCBs and other organic compounds); 2 contaminants for crabs (PCBs

and zinc); 6 contaminants for at least one fish ROC (arsenic, cadmium, copper, total PCBs, TBT, and

vanadium), and 12 contaminants for at least one wildlife ROC (arsenic, cadmium, chromium, cobalt,

copper, lead, mercury, nickel, selenium, total PCBs, zinc, and vanadium). Contaminants in another subset

of COPCs were evaluated only in the uncertainty analysis either because there was uncertainty regarding

their presence at concentrations of concern (i.e., contaminants were never detected in tissue, but reporting

limits (RLs) were above the screening criteria) or because effect-level toxicity information was not

available for them. Otherwise, COPCs were evaluated as discussed below.

7.2.4 Exposure and Effects Assessment

Summaries of exposure pathways determined to be complete for ecological receptors are provided in

Table 16. Toxicity reference values (TRVs) used in the effects assessment were derived from the

scientific literature using survival, growth, and reproduction as assessment endpoints.


60

Figure 11. Conceptual Models for the Ecological Risk Assessment


61

Table 15. Numerical Benthic Sediment Cleanup Objectives and Benthic Cleanup Screening Levels from the Washington State Sediment Management Standards

a

Contaminant Units Basis (dry weight or Organic Carbon)

Benthic Sediment Cleanup Objective

Benthic Cleanup Screening Level

Inorganic Compounds

Arsenic mg/kg Dry 57 93

Cadmium mg/kg Dry 5.1 6.7

Chromium mg/kg Dry 260 270

Copper mg/kg Dry 390 390

Lead mg/kg Dry 450 530

Mercury mg/kg Dry 0.41 0.59

Silver mg/kg Dry 6.1 6.1

Zinc mg/kg Dry 410 960

Organic Compounds

LPAH mg/kg OC 370 780

Naphthalene mg/kg OC 99 170

Acenaphthylene mg/kg OC 66 66

Acenaphthene mg/kg OC 16 57

Fluorene mg/kg OC 23 79

Phenanthrene mg/kg OC 100 480

Anthracene mg/kg OC 220 1200

2-Methylnaphthalene mg/kg OC 38 64

HPAH mg/kg OC 960 5300

Fluoranthene mg/kg OC 160 1200

Pyrene mg/kg OC 1000 1400

Benz(a)anthracene mg/kg OC 110 270

Chrysene mg/kg OC 110 460

Total benzofluoranthenes mg/kg OC 230 450

Benzo(a)pyrene mg/kg OC 99 210

Indeno(1,2,3 c,d)pyrene mg/kg OC 34 88

Dibenzo(a,h)anthracene mg/kg OC 12 33

Benzo(g,h,i)perylene mg/kg OC 31 78

1,2 Dichlorobenzene mg/kg OC 2.3 2.3

1,4 Dichlorobenzene mg/kg OC 3.1 9

1,2,4 Trichlorobenzene mg/kg OC 0.81 1.8

Hexachlorobenzene mg/kg OC 0.38 2.3

Dimethyl phthalate mg/kg OC 53 53

Diethyl phthalate mg/kg OC 61 110

Di-n-butyl phthalate mg/kg OC 220 1700

Butyl benzyl phthalate mg/kg OC 4.9 64

Bis(2-ethylhexyl)phthalate mg/kg OC 47 78

Di-n-octyl phthalate mg/kg OC 58 4500

Dibenzofuran mg/kg OC 15 58

Hexachlorobutadiene mg/kg OC 3.9 6.2


62

Contaminant Units Basis (dry weight or Organic Carbon)

Benthic Sediment Cleanup Objective

Benthic Cleanup Screening Level

n-Nitrosodiphenylamine mg/kg OC 11 11

Total PCBs mg/kg OC 12 65

Phenol µg/kg Dry 420 1200

2-Methylphenol µg/kg Dry 63 63

4-Methylphenol µg/kg Dry 670 670

2,4 Dimethylphenol µg/kg Dry 29 29

Pentachlorophenol µg/kg Dry 360 690

Benzyl alcohol µg/kg Dry 57 73

Benzoic acid µg/kg Dry 650 650

a. SMS biological effects criteria are provided in WAC 173-204-562, Table IV.

Table 16. Assessment Endpoints for Receptors of Concern (ROCs) and Measures of Effect and Exposure

ROC Assessment Endpoint

Assessment Scale Measures of Effect Measures of Exposure

Benthic

Benthic invertebrate community

survival, growth, reproduction

potential exposure area: small exposure areas for individuals

assessment scale: small exposure areas throughout the LDW

SMS and toxicologically based sediment guidelines or TRVs

contaminant concentrations in sediment

water-based TRVs for VOCs

VOC concentrations in porewater

site-specific toxicity tests contaminant concentrations in sediment samples co-located with toxicity test samples

tissue-based TRVs for TBT (excluding imposex in gastropods)

TBT concentrations in sediment samples co-located with benthic invertebrate tissue collection

assessment of imposex in field-collected gastropods

TBT concentrations in sediment samples co-located with gastropod collection

Crabs survival, growth, reproduction

potential exposure area: crabs may forage throughout the LDW

assessment scale: LDW-wide

tissue-based TRVs for decapods

contaminant concentrations in crab tissue collected from four tissue sampling areas located throughout the LDW

Fish

Juvenile chinook salmon

survival and growth

potential exposure area: juvenile salmonids migrate throughout the LDW and forage in shallow areas

assessment scale: intertidal areas throughout the LDW

tissue-based TRVs for contaminants evaluated using a critical tissue-residue approach

contaminant concentrations in juvenile chinook salmon tissue collected from middle and lower segments of the LDW

dietary-based TRVs for contaminants evaluated using a dietary approach

contaminant concentrations in juvenile chinook salmon prey collected from intertidal habitat throughout the LDW, stomach contents collected from juvenile chinook salmon captured throughout the LDW, and sediment collected from intertidal habitats throughout the LDW


63

ROC Assessment Endpoint

Assessment Scale Measures of Effect Measures of Exposure

English sole


potential exposure area: English sole may forage throughout the LDW



contaminant concentrations in English sole tissue collected from four tissue sampling areas located throughout the LDW


contaminant concentrations in English sole prey and sediment collected throughout the LDW

Pacific staghorn sculpin


potential exposure area: sculpin may forage throughout the LDW or small segments of LDW

assessment scale: LDW-wide and four modeling areas


contaminant concentrations in sculpin tissue collected from four tissue sampling areas located throughout the LDW


contaminant concentrations in sculpin prey and sediment collected throughout the LDW and divided into four modeling areas

Wildlife

Great blue heron


potential exposure area: herons may forage in areas of shallow water depths throughout the LDW

assessment scale: LDW-wide intertidal

dietary-based TRVs for birds

contaminant concentrations in heron prey collected throughout the LDW and in sediment collected from intertidal habitats throughout the LDW

Osprey survival, growth, reproduction

potential exposure area: osprey may forage from the top meter of water throughout the LDW



contaminant concentrations in osprey prey collected throughout the LDW and in sediment collected from intertidal habitats throughout the LDW

Spotted sandpiper


potential exposure area: sandpipers predominantly forage within small home range segments of the LDW

assessment scale: three intertidal modeling areas


contaminant concentrations in sandpiper prey and sediment collected from intertidal habitats throughout the LDW

River otter survival, growth, reproduction

potential exposure area: river otters may forage throughout the LDW


dietary-based TRVs for mammals

contaminant concentrations in river otter prey and sediment collected throughout the LDW

Harbor seal survival, growth, reproduction

potential exposure area: harbor seals may forage throughout the LDW


dietary-based TRVs for mammals

contaminant concentrations in harbor seal prey and sediment collected throughout the LDW


64

The following approaches were used in the exposure and effects assessment for the benthic invertebrate

community and crabs:

Risks to the benthic invertebrate community were evaluated by comparing surface sediment

contaminant concentrations and site-specific sediment toxicity test results to the SMS chemical

and biological criteria14

; when these were not available, toxicity-based guidelines from the

Dredged Materials Management Program (DMMP) or TRVs were used. The SMS chemical

criteria for protection of marine benthic invertebrates are based on relationships between

sediment contaminant concentrations and adverse effects on benthic invertebrates (reduced

population size or laboratory toxicity tests showing mortality, reduced growth, or impaired

reproduction) using several hundred samples from the Puget Sound area. The methods used to

develop the SMS criteria are consistent with CERCLA ecological risk assessment methodology.

Risks to the benthic invertebrate community from VOC exposure were evaluated by comparing

VOC concentrations in porewater to TRVs.

Risks to the benthic invertebrate community from TBT exposure were evaluated using results

from a site-specific imposex study with gastropods and by comparing TBT concentrations in

LDW benthic invertebrate tissue to TRVs.

Risks to crabs were evaluated by comparing COPC concentrations in LDW crab tissue to TRVs.

Risks to fish were evaluated using two approaches, depending on the COPC. For TBT and PCBs,

concentrations in LDW fish tissue were compared with concentrations in fish tissue in the scientific

literature associated with adverse effects. For arsenic, cadmium, copper, and vanadium, a dietary

approach was used (concentrations in the diet of LDW fish were compared with concentrations associated

in the scientific literature with adverse effects) because they are metabolically regulated by fish.

Risks to wildlife ROCs were evaluated by comparing estimated COPC concentrations as a dietary dose

for each ROC to dietary doses associated with adverse effects from the scientific literature.

7.2.5 Risk Characterization A hazard quotient (HQ) was used to quantify ecological risk as the ratio of the estimated contaminant

exposure level for the species of concern to the TRV. When the HQ exceeds 1.0, there is a potential for

ecological risk.

HQ = I/TRV

Where:

HQ = Ecological hazard quotient (unitless)

I = Contaminant intake level (mg/kg body weight-day)

TRV = Toxicity reference value (mg/kg body weight-day)

Table 17 compares surface sediment concentrations to benthic SCO criteria. Table 18 shows HQs for

crabs, fish, avian and mammalian wildlife, based on TRVs, which are based on no-observed-adverse-

effects levels (NOAELs) - and lowest-observed-adverse-effects levels (LOAELs). LOAEL-based HQs of

greater than or equal to 1 for PCBs indicated a potential for adverse effects to the benthic invertebrate

community, crabs, spotted sandpiper, and river otter. As discussed below, some LOAEL-based HQs

14

The toxicity tests included: an acute 10-day amphipod (Eohaustorius estuarius) survival test; an acute 48-hr

bivalve larvae (Mytilus galloprovincialis) normal survival test; and a chronic 20-day juvenile polychaete

(Neanthes arenaceodentata) survival and growth test.


65

greater 1 were based on highly uncertain exposure or effects data; in that case, the contaminants were not

designated as COCs. No quantitative ecological risk estimates were calculated for dioxins and furans due

to the lack of tissue data for this compound.

Effects on the benthic invertebrate community were assessed by comparing the contaminant

concentrations in LDW surface sediment and results of site-specific toxicity tests to the SMS criteria (see

text box on page 26). Forty-one contaminants were determined to present risks to benthic invertebrates

because their concentrations in surface sediments exceeded the benthic SCO chemical criteria. For the

subset of 76 samples that were further evaluated using bioassay testing, any sample with a concentration

that exceeded the benthic SCO (or CSL) chemical criteria but did not exceed the biological criteria was

designated as not exceeding the benthic SCO (or CSL). Based on data from the FS, one or more COCs

exceeded benthic SCO criteria in approximately 18%, or 80 acres, of the LDW. In 16 of those 80 acres

(4% of the LDW) the CSL was also exceeded, indicating a higher likelihood of adverse effects. The three

COCs with the most frequent exceedances were PCBs, bis(2-ethylhexyl)phthalate (BEHP), and butyl

benzyl phthalate. For all other contaminants, exceedances occurred in 5% or less of the sediment samples

(Table 17) in locations dispersed through the waterway.

7.2.6 Identification of COCs

Determination of which COPCs were COCs for ecological receptors was based on consideration of the

risk estimates, uncertainties discussed in the ERA, preliminary natural background concentrations, and

expected residual risks following planned early actions in the LDW. The COCs from both the ERA and

the HHRA were the focus for identifying preliminary remediation goals (PRGs) and cleanup areas in the

FS. PCBs were identified as a COC for river otter because they have a higher risk of adverse effects such

as reduced reproductive success from the ingestion of seafood contaminated with PCBs. Estimated

exposures of river otter were greater than the LOAEL by a factor of 2.9 and uncertainties in the risk

estimate were relatively low. In addition, 41 contaminants were selected as COCs for benthic

invertebrates because detected concentrations of these 41 contaminants exceeded the benthic SCO criteria

of the SMS in one or more locations. Other COPCs that exceeded risk thresholds (LOAEL-based HQ

greater than or equal to 1.0) were not selected as COCs because of high uncertainty in the effects or

exposure data, comparisons to preliminary background concentrations, or the expectation of low residual

risk following remediation in EAAs, as discussed in the ERA.

Table 18 provides detailed rationale for the identification of COCs for ecological risk. COPCs that were

not selected as COCs were addressed through focused evaluation in the FS. These contaminants may also

be considered in remedial design for specific areas in or near the LDW and in the post-cleanup

monitoring.

7.2.7 Uncertainty Analysis for the ERA

Uncertainties in the ERA are summarized below.

Estimates of the areal extent of surface sediment with concentrations that exceed SMS criteria are

uncertain because they were estimated by interpolating from individual points at which sediments

were sampled.

Data from field studies (many of them conducted in the Puget Sound region) were not included in

the effects assessment and TRV development because of the difficulty in identifying the cause of

toxicity associated with exposures involving multiple chemical and non-chemical stressors.


66

The potential for adverse effects is uncertain for all exposure concentrations that are above the

NOAEL but below the LOAEL (Table 18) due to lack of data on effects of concentrations

between these values.

Some LOAEL-based TRV values are more uncertain due to uncertainties in the studies reporting

the lowest effects concentrations; for example, for the studies reporting PCB TRVs for English

sole and Pacific staghorn sculpin and for osprey.

Some exposure point concentrations (EPCs) are highly uncertain due to a small number of

samples driving the estimate; for example, the HQ for lead in spotted sandpiper is driven by a

high lead concentration in one benthic invertebrate tissue sample.

No quantitative ecological risk estimates were calculated for dioxins and furans because these compounds

were not analyzed in tissues from the LDW. Thus the level of ecological risk from dioxins and furans is

unknown. However, background-based remedial goals to reduce what is presumed to be unacceptable

human health risk from exposure to dioxins and furans through consumption of seafood from the LDW

will reduce exposures (and therefore risk) to higher-trophic-level ecological receptors, as well.


67

Table 17. Surface Sediment Contaminant Concentrations from FS Dataset, with Comparison to SMS Chemical Criteria for Protection of Benthic Invertebrates

Contaminant

Summary Statistics for Surface Sediments Total Number of Surface Sediment Samples in FS Baseline Dataset Exceedances

Waterway Wide

Minimum Detect

Maximum Detect Meana

Total Samples

Detection Frequency

>Benthic SCO, ≤Benthic CSL, detectedb

>Benthic CSL, detectedb

>Benthic SCO or Benthic CSL, detectedb,c >Benthic SCO

Metals and TBT (mg/kg dw)

Arsenic 1.2 1,100 17 916 94% 5 9 14 1.53%

Cadmium 0.03 120 1.0 894 71% 2 12 14 1.57%

Chromium 4.80 1,680 42 906 100% 1 10 11 1.21%

Copper 5.0 12,000 106 908 100% 0 13 13 1.43%

Lead 2.0 23,000 139 908 100% 2 23 25 2.75%

Mercury 0.015 247 0.53 927 88% 20 30 50 5.39%

Nickel 5.0 910 28 836 100% NA NA NA —

Silver 0.018 270 1.0 875 61% 0 10 10 1.14%

Vanadium 15 150 59 589 100% NA NA NA —

Zinc 16 9,700 194 905 100% 26 19 45 4.97%

Tributyltin as ion 0.28 3,000 90 189 94% NA NA NA —

PAHs (µg/kg dw)

2-Methylnaphalene 0.38 3,300 42 884 19% 1 4 5 0.57%

Acenaphthene 1.0 5,200 65 891 40% 16 4 20 2.24%

Anthracene 1.3 10,000 134 891 73% 2 0 2 0.22%

Benzo(a)anthracene 7.3 8,400 322 891 92% 10 6 16 1.80%

Benzo(a)pyrene 6.5 7,900 309 886 92% 7 5 12 1.35%

Benzo(g,h,i)perylene 6.1 3,800 165 891 86% 10 12 22 2.47%

Total benzofluoranthenes 6.6 17,000 732 885 94% 6 6 12 1.36%

Chrysene 12 7,700 474 891 95% 29 3 32 3.59%

Dibenzo(a,h)anthrecene 1.6 1,500 63 891 56% 18 6 24 2.69%

Dibenzofuran 1.0 4,200 54 889 31% 7 3 10 1.12%

Fluoranthene 18 24,000 889 891 97% 35 12 47 5.27%

Fluorene 0.68 6,800 78 891 48% 11 3 14 1.57%

Indeno(1,2,3-cd)pyrene 6.4 4,300 180 891 90% 16 13 29 3.25%

Naphthalene 3.0 5,300 49 882 21% 0 2 2 0.23%

Phenanthrene 7.1 28,000 429 891 93% 27 3 30 3.37%

Pyrene 19 16,000 723 891 97% 2 6 8 0.90%

Total HPAH 23 85,000 3,809 891 98% 25 6 31 3.48%

Total LPAH 9.1 44,000 696 891 94% 4 3 7 0.79%


68

Contaminant

Summary Statistics for Surface Sediments Total Number of Surface Sediment Samples in FS Baseline Dataset Exceedances

Waterway Wide

Minimum Detect

Maximum Detect Meana

Total Samples

Detection Frequency

>Benthic SCO, ≤Benthic CSL, detectedb

>Benthic CSL, detectedb

>Benthic SCO or Benthic CSL, detectedb,c >Benthic SCO

Phthalates (µg/kg dw)

Bis(2-ethylhexyl) phthalate 5.4 17,000 590 886 79% 46 58 104 11.74%

Butyl benzyl phthalate 2.0 7,100 87 878 54% 80 10 90 10.25%

Dimethyl phthalate 2.0 440 25 878 21% 0 2 2 0.23%

Chlorobenzenes (µg/kg dw)

1,2,4-Trichlorobenzene 1.6 940 19 871 1% 0 2 2 0.23%

1,2-Dichlorobenzene 1.3 670 19 871 2% 0 4 4 0.46%

1,4-Dichlorobenzene 1.5 1,600 23 871 6% 0 4 4 0.46%

Hexachlorobenzene 0.4 95 17 874 5% 4 2 6 0.69%

Other SVOCsd and COCs (µg/kg dw)

2,4-Dimethylphenol 6.1 290 44 869 3% 0 25 25 2.88%

4-Methylphenol 4.8 4,600 44 883 13% 0 4 4 0.45%

Benzoic acid 54 4,500 238 876 13% 0 9 9 1.03%

Benzyl alcohol 8.2 670 49 867 3% 9 7 16 1.85%

Carbazole 3.2 4,200 82 775 55% NA NA NA

n-Nitrosodiphenylamine 6.5 230 27 871 3% 0 2 2 0.23%

Pentachlorophenol 14 14,000 122 840 4% 1 1 2 0.24%

Phenol 10 2,800 91 886 32% 19 6 25 2.82%

Pesticides (µg/kg dw)

Total DDTs 0.72 77,000 462 216 40% NA NA NA —

Total chlordanes 0.20 230 268 216 13% NA NA NA —

Aldrin 0.01 1.6 27 216 2% NA NA NA —

Dieldrin 0.10 280 29 218 4% NA NA NA —

alpha-BHC 0.14 1.8 1.1 207 1% NA NA NA —

beta-BHC 0.09 13 1.2 207 2% NA NA NA —

gamma-BHC 0.05 8.6 27 216 6% NA NA NA

Heptachlor 0.12 5.2 27 216 3% NA NA NA

Heptachlor epoxide 0.47 4.9 2.8 207 2% NA NA NA

Toxaphene 340 6,300 111 205 1% NA NA NA


69

Total PCBs (µg/kg dw)

Total PCBse 2.2 223,000 1,136 1390 94% 336 179 515 37.05%

Source: LDWG (2012) General: Contaminants identified as risk drivers for the benthic invertebrate community (RAO 3) are those with one or more surface sediment samples with exceedances of the SCO. Three additional

contaminants (total DDTs, total chlordanes, and nickel) that do not have SMS criteria were also identified as COCs for the benthic community. a. Calculated mean concentration is the average of concentrations using one-half the reporting limit substitution for non-detected results. b. For non-polar organic compounds, comparisons to SCO and CSL were made using organic carbon-normalized concentrations. If total organic carbon (TOC) in the sample was <0.5% or >4%, dry

weight concentrations were compared to the Apparent Effect Thresholds: (Lowest Apparent Effects Threshold) and Second Lowest Apparent Effects Threshold. Additional discussion can be found at http://www.ecy.wa.gov/programs/tcp/smu/sed_pubs.htm#ApparentEffectsThreshold/. See also Section 15 (Key Terms).

c. Sum of samples exceeding the SCO but not the CSL and samples exceeding the CSL. d. SVOCs — semi-volatile organic compounds e. Total PCB statistics and counts were generated with two outliers excluded (2,900,000 and 230,000 µg/kg dw at RM 2.2).

Table 18. Rationale for Selection of Contaminants as COCs for Ecological Risk

COPC ROC

Maximum NOAEL-

Based HQ

Maximum LOAEL-

Based HQ Additional Considerations COC?

Total PCBs

crabs 10 1.0 Uncertainty in exposure data: whole-body concentrations were estimated Uncertainty in effects data: LOAEL-based HQ was based on a study with Aroclor 1016 and grass shrimp, and NOAEL was estimated using an uncertainty factor; selection of next higher TRV would result in LOAEL-based HQ < 1.0

no

river otter 5.8 2.9 Uncertainty in exposure data: low uncertainty in diet assumptions and home range Uncertainty in effects data: low uncertainty in TRV (growth endpoint in kits)

yes

English sole 4.9 – 25a 0.98 – 5.0a

Uncertainty in exposure data: low uncertainty in tissue concentrations Uncertainty in effects data: high uncertainty in lowest LOAEL TRV because of uncertain statistical significance of the fecundity endpoint for the low dose, a lack of dose-response in the fecundity endpoint, uncertain number of fish used in the experiment, and uncertainties associated with fish handling and maintenance protocols

no


3.8 – 19a 0.76 - 3.8a Same considerations as listed above for English sole no

PCB TEQb

spotted sandpiper –Area 2 (high-quality foraging habitat)

15 1.5

Uncertainty in exposure data: low uncertainty in diet assumptions and home range Uncertainty in effects data: high uncertainty in TRV, which was based on study of reproduction with weekly IP injection; high uncertainty in TEFs; effects data for total PCBs are less uncertain than for PCB TEQs and the LOAEL-based HQ for total PCBs was < 1.0

no

Cadmium

juvenile chinook salmon

5.0 1.0

Uncertainty in exposure data: LOAEL-based HQ < 1.0 if empirical juvenile chinook salmon stomach contents data from the LDW are used to estimate exposure, instead of estimating exposure based on ingestion of benthic invertebrates Uncertainty in effects data: high uncertainty in the lowest TRV because selection of next higher TRV would result in LOAEL-based HQ < 1.0, all salmonid-specific studies for cadmium with NOAELs result in NOAEL-based HQs less than 0.01

no

English sole 6.1 1.2 Uncertainty in exposure data: low uncertainty (LDW-collected benthic invertebrate tissue samples) Uncertainty in effects data: high uncertainty in the lowest TRV because selection of next higher TRV would result in LOAEL-based HQ < 1.0; all other NOAELs and LOAELs were orders of magnitude higher than the selected LOAEL

no


5.2 1.0 Uncertainty in exposure data: low uncertainty (LDW-collected shiner surfperch and benthic invertebrate tissue samples) Uncertainty in effects data: high uncertainty in the lowest TRV because selection of next higher TRV would result in LOAEL-based HQ < 1.0; all other NOAELs and LOAELs were orders of magnitude higher than the selected LOAEL

no

http://www.ecy.wa.gov/programs/tcp/smu/sed_pubs.htm#ApparentEffectsThreshold/


70

COPC ROC

Maximum NOAEL-

Based HQ

Maximum LOAEL-


Chromium

spotted sandpiper –Area 2 (high- and poor-quality foraging habitat)

8.8 1.8

Uncertainty in exposure data: high uncertainty because LOAEL-based HQ would be less than 1.0 if the single anomalously high benthic invertebrate tissue sample from RM 3.0 west was excluded; chromium concentrations in sediment were low in this area Uncertainty in effects data: high uncertainty; only one study with reported effects, and study was unpublished and could not be obtained for review

no

Copper


1.5 1.1

Uncertainty in exposure data: low uncertainty Comparison to natural background: concentration in sediment (Surface Weighted Average Concentration of 57 mg/kg dw) from Area 3 (high- and poor-quality foraging habitat) similar to PSAMP rural Puget Sound concentrations (50 mg/kg dw [90th percentile]) Residual risk: following planned sediment remediation within early action areas, LOAEL-based HQ would be < 1.0

no

Lead


19 5.5

Uncertainty in exposure data: high uncertainty because LOAEL-based HQ would be less than 1.0 if the single anomalously high benthic invertebrate tissue sample from RM 3.0 west was excluded; lead concentrations in sediment were low in this area Uncertainty in effects data: low uncertainty (reproductive endpoint)

no


5.0 1.5 Uncertainty in exposure data: low uncertainty Uncertainty in effects data: low uncertainty (reproductive endpoint) Residual risk: following planned sediment remediation within early action area, LOAEL-based HQ would be < 1.0

Mercury

spotted sandpiper –Area 3 (high- quality foraging habitat)

5.3 1.0 Uncertainty in exposure data: low uncertainty Uncertainty in effects data: low uncertainty (TRV was based on a growth endpoint) Residual risk: following planned sediment remediation within early action area, LOAEL-based HQ would be < 1.0

no

Vanadium

English sole 5.9 1.2

Uncertainty in exposure data: low uncertainty Uncertainty in effects data: high uncertainty in TRV because only one study was available Comparison to natural background: exposure concentration in LDW sediment (SWAC of 58 mg/kg dw) was less than PSAMP rural Puget Sound concentration (64 mg/kg dw [90th percentile])

no


3.2 – 5.9 0.65 – 1.2 Same considerations as listed for English sole above no

spotted sandpiper – all exposure areas

2.0 – 2.7 1.0 – 1.4

Uncertainty in exposure data: low uncertainty Uncertainty in effects data: TRV was based on a 4-week growth endpoint, with uncertainty (two available studies: one with reduced body weight in chickens after 4 weeks and the other with no effect on body weight in mallards after 10 weeks) Comparison to natural background: mean exposure concentrations in sandpiper exposure areas ranged from 49 to 57 mg/kg dw, compared to Puget Sound Ambient Monitoring Program rural Puget Sound background concentration of 64 mg/kg dw (90th percentile)

no

41 SMS contami-nantsc

benthic invertebrates

range of values

range of values

Each of these 41 contaminants had at least one detected exceedance of benthic SCO in baseline surface sediment dataset

yes


71

COPC ROC

Maximum NOAEL-

Based HQ

Maximum LOAEL-


Nickel benthic invertebrates

6.6 2.5

Uncertainty in exposure data: low uncertainty Uncertainty in effects data: medium uncertainty in the TRV (i.e., the ML) because only no-effects data (amphipod mortality and community abundance Apparent Effects Thresholds) were available; no information was available regarding concentrations associated with adverse effects Residual risk: ML was exceeded at four locations in LDW – all within early action areas with planned sediment remediation

no

Total DDTs benthic invertebrates

5.1 2.7

Uncertainty in exposure data: medium uncertainty (i.e., likely interference in pesticide analyses from PCBs) Uncertainty in effects data: medium uncertainty; based on a single study with spiked sediment Residual risk: LOAEL was exceeded at only one location in LDW, location is within early action area with planned sediment remediation

no

Total chlordane

benthic invertebrates

82 48

Uncertainty in exposure data: highly uncertain because all total chlordane concentrations in samples from Phase 2 locations were JN-qualified as a result of probable PCB interference; except one location at RM 2.2, all locations with detected total chlordane concentrations co-occurred with elevated PCB concentrations Uncertainty in effects data: TRV is highly uncertain because it was based on a general Canadian sediment guideline (PEL); this guideline is based mainly on field-collected data with complex mixtures of contaminants Residual risk: LOAEL was exceeded at 14 locations in LDW; all but one of these locations are associated with an early action area with planned sediment remediation

no

Note: HQs for fish are the highest HQs in cases where more than one approach was used. a. LOAEL-based HQs were calculated from a range of effects concentrations reported in Hugla and Thome (1999) because of uncertainty in the LOAEL. The NOAEL TRV range was estimated by

dividing the LOAEL TRV range by an uncertainty factor of 5. Ranges reported for Pacific staghorn sculpin also included the range in exposure estimates for areas smaller than the entire LDW. b. Risk estimates based on TEQs were calculated using only tissue data for dioxin-like PCB congeners because dioxin and furan tissue data were not available. Thus, risks associated with exposure

to all dioxin-like contaminants were likely underestimated; the degree of underestimation is uncertain. c. Arsenic, cadmium, chromium, copper, lead, mercury, silver, zinc, acenaphthene, anthracene, benz(a)anthracene, benzo(a)pyrene, benzo(g,h,i)perylene, chrysene, dibenzo (a,h)anthracene,

fluoranthene, fluorene, indeno(1,2,3,-c,d)pyrene, naphthalene, phenanthrene, pyrene, total benzofluoranthenes, HPAH, LPAH, bis(2-ethylhexyl) phthalate, butyl benzyl phthalate, dimethyl phthalate, 1,2-dichlorobenzene, 1,4-dichlorobenzene, 1,2,4-trichlorobenzene, 2-methylnaphthalene, 4-methylphenol, 2,4-dimethylphenol, benzoic acid, benzyl alcohol, dibenzofuran, hexachlorobenzene, n-nitrosodiphenylamine, pentachlorophenol, phenol, total PCBs.

NOTE: arsenic and total PCBs are also human health contaminants of concern.


72

7.3 Basis for Action

The response action selected in this Record of Decision is necessary to protect public health or welfare or

the environment from actual or threatened releases of pollutants or contaminants at or from this In-

waterway Portion of the Site which may present an imminent and substantial endangerment to public

health or welfare or the environment. A response action is necessary for the In-waterway Portion of the

Site because:

Human Health Risk: The risk of an individual developing cancer or noncarcinogenic effects

related to exposure to contaminants at the In-waterway Portion of the Site exceeds the acceptable

risk range identified in the National Oil and Hazardous Substances Pollution Contingency Plan

(National Contingency Plan; NCP). Specifically, seafood consumption risks and direct contact

HQs for the RME scenarios exceed CERCLA risk thresholds of an excess cancer risk of 1 x 10-4

and a noncancer HQ of 1. MTCA/SMS thresholds are also exceeded, as discussed in Section 7.1.

Ecological Risk: Risks to ecological receptors exceed CERCLA risk thresholds. Forty-one

contaminants were determined to present risks to benthic invertebrates because their

concentrations in surface sediments exceeded the benthic SCO criteria of the SMS. The benthic

SCO criteria are based on studies showing the relationship between contaminant concentrations

and adverse effects to benthic invertebrates. Risks to river otter (based on an analysis of risks to

higher–trophic level species [HTLS]) exceed the LOAEL-based HQ by a factor of 2.9 (Table 18),

and uncertainties in the risk estimate were relatively low. MTCA/SMS thresholds are also

exceeded, as discussed in Section 7.2.


73

8 Remedial Action Objectives

In accordance with the NCP, EPA developed Remedial Action Objectives (RAOs) to describe what the

proposed cleanup is expected to accomplish to protect human health and the environment. The RAOs for

the LDW are based on results of the human health and ecological risk assessments described in Section 7.

RAOs help focus the development and evaluation of remedial alternatives and form the basis for

establishing cleanup levels in the ROD.

8.1 Remedial Action Objectives

The four RAOs established for the LDW are presented below along with a brief summary of how the

Selected Remedy addresses each one:

RAO 1: Reduce risks associated with the consumption of contaminated resident LDW fish and

shellfish by adults and children with the highest potential exposure to protect human health. Risk

will be reduced by reducing sediment and surface water concentrations or bioavailability of PCBs,

arsenic, cPAHs and dioxins/furans, the primary COCs that contribute to the estimated cancer and

noncancer risks from consumption of resident seafood, which will reduce concentrations of these COCs

in tissue. Ongoing source control and the use of seafood consumption advisories and education and

outreach programs will provide additional risk reduction.

RAO 2: Reduce risks from direct contact (skin contact and incidental ingestion) to contaminated

sediments during netfishing, clamming, and beach play to protect human health. Risks will be

reduced by reducing sediment concentrations or bioavailability of PCBs, arsenic, cPAHs, and

dioxins/furans, the primary COCs that contribute to the estimated excess cancer and noncancer risks.

RAO 3: Reduce to protective levels risks to benthic invertebrates from exposure to contaminated

sediments. Risks will be reduced by reducing sediment concentrations of the 41 contaminants listed in

Table 20 to the chemical or biological benthic SCO.

RAO 4: Reduce to protective levels risks to crabs, fish, birds, and mammals from exposure to

contaminated sediment, surface water, and prey. Risks will be reduced by reducing sediment and

surface water PCB concentrations or bioavailability, which will reduce PCB concentrations in tissue.

Addressing risks to river otters due to consumption of PCB-contaminated seafood, along with addressing

risks associated with RAOs 1 – 3, will also address risks to other ecological receptors.

8.2 Cleanup Levels, ARARs and Target Tissue Concentrations

This section describes the selected cleanup levels (see Section 8.2.1), ARARs (see Section 8.2.2), and

target tissue concentrations (see Section 8.2.3) for the in-waterway cleanup and key factors that formed

the basis for each. The selected cleanup levels are contaminant concentrations that will be used to

measure the success of the cleanup alternatives in meeting the RAOs. Cleanup levels are based on

applicable or relevant and appropriate requirements (ARARs), which provide minimum legal standards,

and other information such as toxicity information from the HHRA and ERA.


74

8.2.1 Cleanup Levels

Table 19 lists sediment cleanup levels for RAOs 1, 2, and 4, and Table 20 lists sediment cleanup levels

for RAO 3. Sediment cleanup levels for contaminants for RAO 3 are point-based and applicable to any

sample location; for the other RAOs, cleanup levels are applied to a specific area (see Table 19). Benthic

cleanup levels are based on the benthic SCO in the SMS (WAC 173-204-562). For RAO 3, the SCO

numerical chemical criteria can be overridden by the SCO biological criteria (see text box "What are the

Sediment Management Standards?" on page 26) unless they are co-located with exceedances of remedial

action levels (RALs) associated with human health COCs, which are also point-based. Exceedances of

RALs for human health COCs cannot be overridden by toxicity testing.

Table 19. Cleanup Levels for PCBs, Arsenic, cPAHs, and Dioxins/Furans in Sediment for Human Health and Ecological COCs (RAOs 1, 2 and 4)

COC

Cleanup Levels Application Area and Depth

RAO 1: Human Seafood Consumption

RAO 2: Human Direct Contact

RAO 4: Ecological (River Otter)

Basis for Cleanup Levelsa

Spatial Scale of Applicationb

Spatial Compliance Measuree

Compliance Depthb

PCBs (µg/kg dw)

2 1,300 128

background (RAO 1) RBTC (RAO 2) RBTC (RAO 4)

LDW-wide UCL95 0 – 10 cm

NA 500 NA RBTC All Clamming Areasc

UCL95 0 – 45 cm

NA 1,700 NA RBTC Individual Beachesd

UCL95 0 – 45 cm

Arsenic (mg/kg dw)

NA 7 NA background LDW-wide UCL95 0 – 10 cm

NA 7 NA background All Clamming Areasc

UCL95 0 – 45 cm

NA 7 NA background Individual Beachesd

UCL95 0 – 45 cm

cPAH (µg TEQ/kg dw)

NA 380 NA RBTC LDW-wide UCL95 0 – 10 cm


UCL95 0 – 45 cm

NA 90 NA RBTC Individual Beachesd

UCL95 0 – 45 cm

Dioxins/Furans (ng TEQ/kg dw)

2 37 NA background (RAO 1) RBTC (RAO 2)

LDW-wide UCL95 0 – 10 cm


UCL95 0 – 45 cm

NA 28 NA RBTC Individual Beachesd

UCL95 0 – 45 cm

NOTE: where there are multiple cleanup levels for a cleanup area, the lowest cleanup level is shown in bold. a. Background – see Table 3 and Section 5.3.4.1; RBTC – Risk-based threshold concentration (based on 1 in 1,000,000 excess cancer risk

or HQ of 1) b. In intertidal areas including beaches used for recreation and clamming, human-health direct contact cleanup levels (for PCBs, arsenic,

cPAHs, and dioxins/furans) must be met in the top 45 cm because in intertidal areas exposure to sediments at depth is more likely through digging or other disturbances. Human health cleanup levels for RAO 1 (seafood consumption) and ecological cleanup levels must be met in surface sediments (top 10 cm). In subtidal areas, cleanup levels for all COCs must be met in surface sediments (top 10 cm).

c. Clamming areas are identified in Figure 6. d. Beach play areas are identified in Figure 6. e. The UCL 95 is the upper confidence limit on the mean. The determination of compliance with RAOs 1, 2 and 4 cleanup levels will be made

by one of two methods: 1) comparison of the UCL 95 of LDW data with the RBTC or background-based cleanup level, or 2) for background-based cleanup levels, a statistical comparison of the distribution of LDW data to the OSV BOLD study background dataset (USACE et al. 2009) may be used. In either case, testing will use an alpha level of 0.05 and a beta level of 0.10. For details, see ProUCL technical manual (EPA 2013b) or most current version). For either method, a sufficient number of samples must be collected to assure statistical power for the test.


75

Table 20. Sediment Cleanup Levels for Ecological (Benthic Invertebrate) COCs for RAO 3a

Benthic COC Cleanup Level for RAO 3a Benthic COC Cleanup Level for RAO 3a

Metals, (mg/kg dw)c OC-normalized Organic Compounds (continued) (mg/kg OC)

Arsenic 57

Total PCBs 12

Cadmium 5.1 Benzo(g,h,i)perylene 31

Chromium 260 Chrysene 110

Copper 390 Dibenz(a,h)anthracene 12

Lead 450 Indeno(1,2,3-cd)pyrene 34

Mercury 0.41 Fluoranthene 160

Silver 6.1 Fluorene 23

Zinc 410 Naphthalene 99

Dry Weight Basis Organic Compounds, (µg/kg dw) Phenanthrene 100

4-methylphenol 670 Pyrene 1,000

2,4-dimethylphenol 29 HPAH 960

Benzoic acid 650 LPAH 370

Benzyl alcohol 57 Bis(2-ethylhexyl)phthalate 47

Pentachlorophenol 360 Butyl benzyl phthalate 4.9

Phenol 420 Dimethyl phthalate 53

1,2-dichlorobenzene 2.3

OC-normalized Organic Compounds, (mg/kg OC)b 1,4-dichlorobenzene 3.1

Acenaphthene 16 1,2,4-trichlorobenzene 0.81

Anthracene 220 2-methylnaphthalene 38

Benzo(a)pyrene 99 Dibenzofuran 15

Benz(a)anthracene 110 Hexachlorobenzene 0.38

Total benzofluoranthenes 230 n-Nitrosodiphenylamine 11

a. Cleanup Levels for RAO 3 are based on the benthic SCO chemical criteria in the SMS (WAC 173-204-562). Benthic SCO biological criteria (WAC 173-204-562, Table IV) may be used to override benthic SCO chemical criteria where human health-based RALs are not also exceeded.

b. PCBs and arsenic are also human health COCs; see Table 19.

No sediment cleanup levels were identified for arsenic or cPAHs for the human health seafood

consumption pathway (RAO 1). Seafood consumption excess cancer risks for these two COCs were

largely attributable to eating clams. However, data collected during the RI/FS showed little relationship

between concentrations of arsenic or cPAH in sediment and their concentrations in clam tissue. EPA will

define the sediment cleanup footprint based on other cleanup levels, then use the clam target tissue levels

(Section 8.2.3) to measure reduction in arsenic and cPAH concentrations in clams. Research will be

conducted during the remedial design phase to study the relationships between sediment concentrations

for arsenic and cPAHs and concentrations in clam tissue and methods to reduce concentrations of these

contaminants in clams. If EPA determines, based on these studies, that additional remedial action is

needed to reduce clam tissue arsenic and cPAH concentrations for the purpose of achieving RAO 1, EPA

will document and select those actions in a future decision document.


76

The sediment cleanup levels for PCBs and dioxins/furans (RAO 1) and for arsenic (RAO 2) are set at

natural background consistent with the SCO for human health risks (HH SCO). Modeling conducted

during the RI/FS could not predict that long term LDW COC concentrations would achieve natural

background. This is because the concentrations of these contaminants in incoming sediments (suspended

solids) from the Green/Duwamish River are currently higher than natural background and current

practical limitations on control of sources within the LDW and Green/Duwamish River drainage basins

may not allow sufficient future reductions in these incoming concentrations. The term cleanup objective

was used in the FS to mean the PRG or as close as practicable to the PRG (sediment PRGs in the FS and

Proposed Plan are cleanup levels in the ROD). This ROD uses the term “FS cleanup objective” when

referring to the term as it was used in the FS to distinguish it from the new term SCO in the 2013 SMS.

For the purposes of comparing alternative remedies, the lowest model-predicted concentration was used

as a surrogate for “as close as practicable to the PRG” when the PRG was not predicted to be achieved

within a 45-year period.

These long-term COC concentrations predicted by the model are highly uncertain. As discussed in the FS

(LDWG 2012a), concentrations of COCs coming in to the LDW from upstream and lateral sources vary

over time and are difficult to predict; therefore, the values used to represent these COC concentrations,

used as model inputs, are uncertain. In particular, the data used to estimate Green/Duwamish River

surface water and sediment inputs to the RI/FS models were relatively sparse and highly variable. In

addition, it is difficult to predict what concentrations in upstream and lateral-source sediments will be

many years in the future. High and low bounds on these inputs were evaluated in the FS to portray model

sensitivity. For example, RI/FS models predict that all alternatives will reduce PCB concentrations in

LDW sediments to approximately 40 – 45 µg/kg in 40 years using mid-range model input parameters

(Table 5). In contrast, the sensitivity analysis indicates that future PCB sediment concentrations could

range from 9 – 100 µg/kg. The great majority of this range is due to varying assumptions about incoming

suspended sediment concentrations. Ecology and King County are currently conducting studies to refine

estimates of contaminant inputs from the Green/Duwamish River, and to better understand upstream

sources of contamination. Ecology in coordination with EPA will use this information to further assess

upstream source control. EPA is retaining natural background, along with the risk-based values (RBTCs),

as the basis for cleanup levels for LDW sediments.

8.2.2 ARARs ARARs are legally applicable or relevant and appropriate substantive (as opposed to administrative)

standards, requirements, criteria, or limitations under any federal environmental law, or promulgated

under any state environmental or facility siting law that is more stringent than under federal law. This

section discusses MTCA and surface water quality requirements; these ARARs are also discussed in

Sections 10.1.2 and 14.2, and a complete list of ARARs is in Table 26.

8.2.2.1 Sediment Quality ARARs

The most significant ARARs for developing cleanup levels during the RI/FS and for the Proposed Plan

for the In-waterway Portion of the Site were in MTCA and its rules in WAC 173-340 for Washington

cleanup sites generally, and the SMS rules for sediment cleanups in WAC 173-204, which are referred to

in the MTCA general cleanup rules (WAC 173-340-760). Major portions of the SMS were revised in

September 2013, after the Proposed Plan was issued, in part to update sediment cleanup requirements in

Part V (Sediment Cleanup Standards) of the SMS and harmonize Part V requirements with the


77

requirements in MTCA. The 1991 SMS was promulgated under several authorities including both MTCA

and the state Water Pollution Control Act. However, Part V of the 2013 SMS was promulgated solely

under MTCA. See “What are the Sediment Management Standards?” on page 26 for a summary of the

2013 SMS. As a matter of substance, the MTCA and SMS-based sediment PRGs set forth in the Proposed

Plan using the 1991 SMS remain unchanged as cleanup levels in the ROD, though the method for

deriving them (applying the substantive requirements of the 2013 SMS) is different, as explained below.

This section describes the derivation of the cleanup levels in this ROD in terms of the revised SMS rules.

Sediment cleanup levels for RAOs 1 and 2 (for protection of human health) are calculated at the SCO

level – risk-based threshold concentrations (RBTCs) of 1 x 10-6

excess cancer risk for individual

carcinogens, 1 x 10-5

excess cancer risk cumulatively for multiple carcinogens, and noncancer HQ or HI

of 1, consistent with the NCP and as required by the revised SMS (WAC 173-204-560 and 561). In

accordance with the SMS, where RBTCs at SCO levels are more stringent than background levels, the

SCO-based cleanup levels are set at the natural background level (see Section 5.3.4.1)15

.

Similarly, consistent with the revised SMS (WAC 173-204-562), cleanup levels associated with RAO 3

(protection of benthic invertebrates) are based on the SCO for the protection of benthic invertebrates

(benthic SCO) of the SMS which are defined by chemical and biological criteria for specific hazardous

substances as explained in Section 5.3.1.1. The benthic SCO chemical and biological criteria are the same

as the 1991 SMS Sediment Quality Standards criteria used in the FS and Proposed Plan. EPA also

considered risks to higher-trophic-level species (HTLS) (WAC 173-204-564) in setting a PCB cleanup

level for river otter (RAO 4). Cleanup levels for the protection of human health and benthic invertebrates

are also protective of HTLS.

The 2013 SMS (WAC 173-204-560) requires initial establishment of cleanup levels at the SCO level, but

allows for the cleanup levels to be adjusted upward to CSL levels when it is not technically possible to

achieve SCO levels, or if meeting the SCO will have a net adverse impact on the aquatic environment.

CSL risk-based cleanup levels are the most stringent of the following: 1) for human health, an excess

cancer risk of 1 x 10-5

for individual carcinogens and for multiple carcinogens cumulatively, and a

noncancer HQ or HI of 1; 2) for risks to benthic invertebrates, chemical and biological criteria defined in

WAC 173-204-562 (which are the same as the CSL criteria in the 1991 SMS); and 3) for risks to HTLS,

the same no-observed-adverse-effects threshold as the SCO per WAC 173-204-564. The CSL is the

highest of the risk based concentration, PQL, or regional background (a new term created by the 2013

SMS). There is insufficient information at this time to determine whether or not it is technically possible

to achieve the SCO-based cleanup levels selected in this ROD, for the reasons discussed in Section 8.2.1.

In addition, neither EPA nor Ecology has established regional background for the LDW.

If long-term monitoring data and trends indicate that some cleanup levels or other ARARs cannot be met,

EPA will determine whether further remedial action could practicably achieve the ARAR. If EPA

concludes that an ARAR cannot be practicably achieved, EPA will waive the ARAR on the basis of

technical impracticability (TI) in a future decision document (ROD Amendment or ESD). For SMS SCO-

based ARARs, EPA will first consider whether the criteria in the SMS for adjusting cleanup levels from

15

The SMS also allows upward adjustment for cleanup levels that are below practical quantitation limits (PQLs);

however, this is not applicable for the LDW, where natural background- and risk-based cleanup levels are higher

than PQLs.


78

the SCO to the CSL (including regional background) can be met, as discussed above. If these criteria can

be met, EPA will evaluate adjusting the relevant sediment cleanup levels upward to regional background

or other CSL-based levels described in the SMS.

8.2.2.2 Surface Water Quality ARARs

Surface water quality ARARs consist of applicable promulgated state water quality standards and, in

accordance with Section 121(d)(2)(A)(ii) and (B)(i) of CERCLA, federal recommended Clean Water Act

Section 304(a) Ambient Water Quality Criteria (AWQC) guidance values where they are relevant and

appropriate. The AWQC for human health include values to protect for consumption of organisms only,

and those to protect for consumption of organisms and water. For the LDW, the relevant and appropriate

AWQC for the protection of human health are those established for the consumption of organisms only

because surface water within the In-waterway Portion of the Site is not a source of consumable water. The

AWQC also include acute and chronic criteria values for the protection of aquatic life, including benthic

organisms. State standards in Washington include those standards promulgated in WAC 173-201A and,

for protection of human health, EPA’s 1992 promulgated National Toxics Rule (NTR) standards (see

Table 26 for legal citations). Consistent with Section 121(d) of CERCLA, the NCP, and MTCA at WAC

173-340-730(3)(b), ARARs are the most stringent of values from WAC 173-201A, NTR, and relevant

and appropriate AWQC.

Surface water will not be directly remediated but will be improved by implementation of the Selected

Remedy and by source control to be implemented as discussed in Section 4.2. Surface water is a key

exposure pathway for consumption of aquatic organisms by humans or wildlife. Surface water quality

data will be compared to these ARAR values to measure progress towards achieving RAOs 1 and 4, and

evaluated as discussed in Section 8.2.2.1.

8.2.3 Fish and Shellfish Target Tissue Concentrations

EPA has established fish and shellfish target tissue concentrations to measure progress toward achieving

RAOs 1 and 4. Controlling sources of contamination to the LDW along with remediating contaminated

sediments will reduce COC concentrations in surface water and in fish and shellfish tissue in addition to

reducing COC concentrations in sediment. Table 21 lists resident fish and shellfish (crab and clam) target

tissue concentrations for RAO 1. They are based on the higher of: the RBTC at 1 x 10-6

excess cancer risk

or HQ of 1 for the adult Tribal RME scenario; or the current concentrations in non-urban (background)

Puget Sound data. Fish and shellfish target tissue concentrations have been developed consistent with the

criteria for developing the sediment cleanup levels (which are based on the 2013 SMS) to measure

protectiveness for humans, including sensitive subpopulations.

Target tissue concentrations are not cleanup levels; they will be used for informational purposes to assess

ongoing risks to people who may consume resident LDW fish and shellfish. Tissue monitoring data will

also inform the content or degree of any potential future fish advisories, other ICs intended to minimize

risk to the LDW fishing community, or other response actions that may be identified in a ROD

Amendment or ESD.

As discussed in Section 5.3.4.2, fish and shellfish target tissue concentrations based on background data

are uncertain because they were developed with a limited dataset. Additional fish and shellfish

background data will be collected during the remedial design phase to increase understanding of non-

urban tissue concentrations of the human health COCs.


79

The FS used the BCM to predict post-cleanup sediment concentrations for various alternatives and the

FWM to predict fish and shellfish tissue PCB concentration associated with these changed sediment

concentrations. Together, these models predicted that background-based fish and shellfish target tissue

PCB concentrations will not be met in the long term due to the same assumptions subject to the same

uncertainties described in the last paragraph of Section 8.2.1. Further, while the same approach was used

to develop target tissue concentrations and sediment cleanup levels, it is not known whether achievement

of sediment cleanup levels would result in the achievement of target tissue levels. Sediment and tissue

background data were not collected concurrently or at the same locations, and food web relationships in

the Puget Sound bays where the natural background samples were taken are likely to be different than in

the Duwamish estuary.

Table 21. LDW Resident Fish and Shellfish Target Tissue Concentrations

Species Group and Tissue Type Speciesa,b, Target

Concentration

Source of Target

Concentration c

PCBs (μg/kg ww)

Benthic fish, fillet English sole 12 Non-urban background

Pelagic fish, whole body Perch 1.8 Species-specific RBTCd

Crab, edible meat Dungeness crab 1.1 Non-urban background

Crab, whole body Dungeness crab 9.1 Non-urban background

Clams Eastern softshell clam 0.42 Non-urban background

Inorganic arsenic (mg/kg ww)

Clamse Eastern softshell clam 0.09 Non-urban background

cPAH TEQ (μg/kg ww)

Clamse Eastern softshell clam 0.24 Species-specific RBTCd

Dioxin/furan TEQ (ng/kg ww)

Benthic fish, whole body English sole 0.35 Non-urban background

Crab, edible meat Dungeness crab 0.53 Non-urban background

Crab, whole body Dungeness crab 2.0 Non-urban background

Clams Eastern softshell clam 0.71 Non-urban background

a Substitutions of similar species may be made if sufficient numbers of the species listed here are not available. b. For non-urban background statistics, see also Table 4. Non-urban background is based on UCL95. c. The statistic used to compare site data to target tissue concentrations will be based on the UCL95 for each compound listed for fish and

crabs collected throughout the waterway; and each compound for clams collected across all clamming areas in the waterway. d. Species-specific RBTCs were used to determine target concentration when RBTCs exceed background, or background data were not

available. e. Only clam tissue values are shown for inorganic arsenic and cPAH TEQ because most of the risk associated with these COCs was

associated with consumption of clams.


80

9 Description of Alternatives

Remedial alternatives for the LDW were developed to meet the requirements of CERCLA and its

regulations, the NCP, which include ARARs such as MTCA and its regulations including the SMS. The

NCP requires that a range of remedial alternatives be evaluated to provide protection of human health and

the environment primarily by preventing or controlling exposure to hazardous substances, pollutants, or

contaminants within a site. The development and analysis of the remedial alternatives form the basis for

EPA's selection of the Selected Remedy and are discussed below.

9.1 Framework for Developing Alternatives

EPA considered several factors in developing remedial alternatives, including: the levels of COCs in

surface and subsurface sediments, the likelihood of humans or aquatic organisms coming into contact

with contaminated sediments, the likelihood that sediment disturbances (many of which can result from

ordinary use of the waterway) might expose contamination in the future, and the potential for

contaminated sediments to be covered by incoming cleaner sediments and therefore pose less risk. EPA

also considered use of the waterway by people and aquatic organisms, as discussed in Section 6. To

support the development of alternatives, EPA developed three criteria: 1) Remedial Action Levels (RALs)

(described below); 2) cleanup levels (described in Section 8.2.1); and 3) Recovery Categories, described

below.

Remedial Action Levels (RALs) are contaminant-specific sediment concentrations that will be used to

identify specific areas that require active remediation (dredging, capping, enhanced natural recovery

[ENR], or a combination thereof), taking into consideration the human health and ecological risk

reduction that could be achieved by the different remedial technologies. These RALs vary by alternative

and are set by EPA so that, in each area, cleanup levels will be met either immediately after construction,

or in the long term after natural recovery, to the extent practicable given the uncertainties discussed in

Section 10.1. The sediment RALs in this ROD are equal to or higher than the sediment cleanup levels for

each COC and are used only to delineate the Site into areas where different remedial technologies would

be used. The use and application of RALs does not affect or alter the requirement to achieve cleanup

levels.

A number of remedial alternatives were developed and are presented in this ROD. Table 22 provides a

summary of the alternatives, which are describe in detail in Section 9.4. As shown in Table 22, each

alternative has its own set of sediment RALs, from higher concentrations (less active cleanup) to lower

concentrations (more active cleanup). Alternatives vary with regard to: which remedial technologies are

used; risk reduction projected to be achieved over time; and the extent to which they rely upon natural

recovery to reduce contaminant concentrations.

Different RALs were established for surface and subsurface sediments, intertidal and subtidal sediments,

and Recovery Category areas.

Contaminant-specific RALs for surface sediments are compared to contaminant concentrations averaged

over the top 10 cm (4 in) of sediments. Consistent with the SMS, the top 10 cm represents the biologically

active zone where most of the benthic invertebrates reside. For subsurface sediments in intertidal areas

(shallower than -4 ft MLLW), certain RALs (identified as intertidal RALs in Table 22) are also compared

to the contaminant concentration averaged over the top 45 cm (1.5 ft). For subsurface sediments in


81

intertidal and subtidal areas with a higher potential for erosion or scour (see Recovery Category 1

description below), RALs are also compared to the contaminant concentration in the top 60 cm (2 ft).16

Where concentrations exceed the RALs, active remediation technologies are selected based on technology

applicability for conditions in the waterway, including relative abilities of the technologies to address

contamination given potential for scour, potential for human contact, site constraints (such as docks and

navigation), and recovery potential. RALs are applied at each discrete sampling location, not as averaged

values applied over the surface area of the waterway sediments. While RALs were used in the FS to

identify areas for each altnernative requiring active remediation, the areas of active remediation will be

further defined through sampling conducted during remedial design. Sampling results will be used to

determine the areal extent and depth of contamination to be addressed by cleanup of the EAAs.

Cleanup Levels are described in Section 8. For each alternative, the projected short-term and long-term

sediment COC concentrations after implementation (developed using the RALs and Recovery Categories)

are compared to cleanup levels to estimate that alternative’s protectiveness and compliance with ARARs.

Recovery Categories were used to assign remedial technologies to specific areas based on information

about the potential for sediment contaminant concentrations to be reduced through natural recovery or for

subsurface contamination to be exposed at the surface due to erosion or scour. Based on data collected

and modeling performed in the RI/FS, three Recovery Categories were developed as shown in Table 23.

The spatial extent of the areas assigned to each of these three categories in the FS is shown in Figure 12.

The use of Recovery Categories allows for more aggressive remedial technologies (such as capping and

dredging) in areas with less potential for natural recovery and a higher likelihood of scour or other

disturbance, and less aggressive remedial technologies (such as ENR and MNR) in areas where recovery

is predicted to occur more readily and disturbance is less likely.

9.2 Summary of Remedial Alternatives

Using the framework described above, along with other criteria such as maintaining sufficient water

depths for human use and habitat areas, 12 remedial action alternatives were developed in the FS using

varying combinations of technologies as described below. The FS alternatives include one no-further-

action alternative (Alternative 1), seven removal-emphasis alternatives (“R” Alternatives 2R, 2R-

Contained Aquatic Disposal (CAD), 3R, 4R, 5R, 5R-Treatment, and 6R) and four combined technology

alternatives (“C” Alternatives 3C, 4C, 5C, and 6C). FS Alternative 5C was further modified to include

additional remedial elements as described in LDWG 2012b and LDWG 2013; the modified alternative is

called 5C Plus. The Selected Remedy is based on Alternative 5C Plus; however, some additional

modifications were made after consideration of public comments on the Proposed Plan. These

modifications are summarized in this section and described in more detail in Section 12. An

approximation of the locations of sediments addressed by the cleanup alternatives is shown in Figure 13.

16

The Selected Remedy, developed after completion of the FS and described in Section 13, adds PCB RALs for

the top 60 cm (2 ft) in Recovery Category 2 and 3 areas, and adds a requirement that all shoaled areas in the

navigation channel that exceed RALs must be dredged.


82

[This page is intentionally blank for double sided printing.]


83

Table 22. Remedial Alternatives and Associated Remedial Technologies, Remedial Action Levels, and Actively Remediated Acres

Remedial Alternatives and Technologiesa

Remedial Action Levelsa Actively

Remediated Area

(Acres)

PCBs

(mg/kg OC)b

Arsenic

(mg/kg dw)

cPAHs

(µg TEQ/kg dw)

Dioxins/ Furans

(ng TEQ/kg dw)

Benthic SMS

(41 Contaminants)b

Alternative 1 No Further Action after removal or capping of Early Action Areas n/a n/a n/a n/a n/a 29 acres

Alternative 2 (2R) – dredge emphasis with upland disposal/MNR

Alternative 2 with CAD (2R-CAD) – dredge emphasis with contained aquatic disposal/MNR

65 to 110 (LDW-wide);

10-yr post-construction target: 65c

93 5,500 50 CSL to 3 × CSL

10-yr post-construction target: CSL

32 acres

Alternative 3 removal (3R) – dredge emphasis with upland disposal/MNR

Alternative 3 combined technologies (3C) – ENR/ in situ / cap/ MNR where appropriate, otherwise dredge with upland disposal

65 (LDW-wide) 93 (LDW-wide)

28 (intertidal)

3,800 (LDW-wide)

900 (intertidal)

35 (LDW-wide)

28 (intertidal)

CSL (biological or chemical) 58 acres

Alternative 4 removal (4R) – dredge emphasis with upland disposal/MNR

Alternative 4 combined technologies (4C) – ENR/ in situ / cap/ MNR where appropriate, otherwise dredge with upland disposal

12 to 35 (LDW-wide)

10-yr post-const. target: 12c

57 (LDW-wide)

28 (intertidal)

1,000 (LDW-wide)

900 (intertidal)

25 (site-wide)

28 (intertidal)

SCO to CSL

10-yr post-const. target: SCO

107 acres

Alternative 5 removal (5R) – dredge emphasis with upland disposal

Alternative 5 removal with treatment (5R-T) – dredge with soil washing treatment and disposal/re-use

Alternative 5 combined technologies (5C) – ENR/ in situ / cap where appropriate, otherwise dredge with upland disposal


28 (intertidal)

1,000 (LDW-wide)

900 (intertidal)

25 (LDW-wide)

28 (intertidal)

SCO (biological or chemical) 157 acres

Alternative 6 removal (6R) – dredge emphasis with upland disposal

Alternative 6 combined technologies (6C) – ENR/ in situ / cap where appropriate, otherwise dredge with upland disposal


28 (intertidal)

1,000 (LDW-wide)

900 (intertidal)

15 (LDW-wide)

28 (intertidal)

SCO (biological or chemical) 302 acres

Selected Remedy (5C Plus) – ENR/ in situ / cap where appropriate; otherwise, dredge with upland disposale 12 (LDW-wide)

65 (intertidal)

195 (subtidal subsurface)

57 (LDW-wide)

28 (intertidal)

1,000 (LDW-wide)

900 (intertidal)

25 (LDW-wide)

28 (intertidal)

2 X SCO chemical criteria d with

10-year post-construction target to meet

SCO

177 acres

a. Areas where remedial action levels (RALs) are applied are as follows: LDW-wide RALs, in the upper 10 cm of sediment throughout the LDW and in the upper 60 cm in potential scour areas (i.e., Recovery Category 1 areas). In intertidal areas, intertidal RALs are applied in the upper 45 cm of sediment (above -4 ft MLLW). Alternative 5C Plus added an intertidal PCB RAL of 65 mg/kg OC in the top 45 cm in intertidal areas, and added a subtidal PCB RAL of 195 mg/kg OC for the top 60 cm in areas of potential vessel scour within Recovery Category 2 and 3 areas. These additional potential vessel scour areas comprise: north of the 1st Avenue South bridge (located at approximately RM 2) in water depths from -4 to -24 ft MLLW, and south of the 1st Avenue S bridge, in water depths from -4 to -18 ft MLLW.

b. See Table 15 for SCO and CSL values. PCB RALs are normalized to organic carbon (OC) for consistency with the SMS, and because the organic content of sediments affects the bioavailability and toxicity of PCBs. The terms SCO and CSL in this table mean the benthic SCO and CSL; SCO is equivalent to the term "SQS" used in the RI/FS and Proposed Plan. Lower human health-based RALs for PCBs and arsenic in this table take precedence over benthic SCO or CSL values.

c. The RALs for SMS contaminants (except arsenic) are a range for Alternatives 2 and 4. The upper RALs are used where conditions for recovery are predicted to be more favorable (Recovery Category 3 areas); the lower RALs are used where conditions for recovery are predicted to be limited or less certain (Recovery Category 1 or 2 areas), or where the BCM does not predict recovery to the 10-yr post-construction target concentration.

d. The Alternative 5C Plus RAL of "2 X SQS not to exceed CSL" in the Proposed Plan is modified in the Selected Remedy to "2 X benthic SCO", see Section 12.

e. The Selected Remedy includes additional requirements to address contaminated shoals in the navigation channel, see Sections 12 and 13.

Table 23. Criteria for Assigning Recovery Categoriesa

Criteria

Recovery Categories

Category 1

Recovery Presumed to be Limited

Category 2

Recovery Less Certain

Category 3

Predicted to Recover

Physical Criteria

Physical

Conditions

Vessel scour Observed vessel scour No observed vessel scour

Berthing areas Berthing areas with vessel scour Berthing areas without vessel scour Not in a berthing area

Sediment Transport

Model

STM-predicted 100-year high-flow scour (depth in cm) > 10 cm < 10 cm

STM-derived net sedimentation using average flow

conditions Net scour Net sedimentation

Rules for applying criteria If an area is in Category 1 for any one criterion, that area is

designated Category 1

If conditions in an area meet a mixture of Category 2 and 3 criteria,

that area is designated Category 2

An area is designated Category 3 only if it meets all Category 3

criteria

Empirical Contaminant Trend Criteria – used on a case-by-case basis to adjust recovery categories that would have been assigned based on physical criteria

Empirical

Contaminant Trend

Criteria

Resampled surface sediment locations Increasing PCBs or increasing concentrations of other detected

COCs that exceed the SCO ( > 50% increase)

Equilibrium and mixed (increases and decreases) results (for COCs

that exceed the SCO)

Decreasing concentrations ( > 50% decrease) or mixed results

(decreases and equilibrium) Sediment cores

(top 2 sample intervals in upper 60 cm)

a. Recovery categories were not assigned to the Early Action Areas, for which remediation should be complete by the time of the remedial actions addressed in this ROD. At the time of the remedial design, EPA will consider assignment of categories to these areas based upon the logic in this table; this information will inform long term monitoring decisions.


84



85

Figure 12. Recovery Category Areas


86

Figure 13. Areas Addressed by LDW Cleanup Alternatives in the FS


87

9.3 Technologies Common to All Remedial Alternatives

The remedial cleanup technologies described below that were used to develop remedial action alternatives

to address contamination in the LDW include dredging and excavation, capping, treatment, enhanced

natural recovery (ENR), and monitored natural recovery (MNR).

The “no action” alternative would use no remedial technologies (although it does include long-term

monitoring). All alternatives would be implemented after completion of cleanup in the Early Action

Areas (29 acres) along with sufficient source control to minimize recontamination.

The engineered remedial technologies are as follows:

Dredging and Excavation – Removal of sediments through dredging or excavation is

incorporated into all remedial alternatives except the no action alternative.

Sediment disposal – All alternatives include disposal of dredged or excavated materials at an off-

site upland permitted facility. Alternative 2R-CAD also includes disposal of contaminated

sediments in a contained aquatic disposal (CAD) site within the LDW. Alternative 5R-Treatment

includes treatment of dredged sediments prior to disposal.

Capping – Many alternatives include capping of contaminated sediments in areas where water

depth is sufficient for a cap. Engineered sediment caps are constructed by placing clean sand,

gravel, and rock on contaminated sediments to provide physical and chemical isolation of

contaminants. Cap thickness will be determined during remedial design; in the FS , caps were

assumed to be 3 ft thick. In habitat areas (see Section 13.2.1.1), the uppermost layers of caps will

use suitable habitat materials. Other materials, such as activated carbon or other contaminant-

sequestering agents, may be used to reduce the potential for contaminants to migrate through the

cap.

Enhanced Natural Recovery (ENR) – Many alternatives include Enhanced Natural Recovery of

contaminated sediments. ENR refers to the placement of a thin layer (approximately 6 to 9

inches) of clean sand (or other suitable habitat materials) on sediments, which immediately

provides a new surface substrate of clean sediments. This cleaner material then mixes with the

underlying contaminated material, through mechanisms such as bioturbation. ENR reduces

contaminant concentrations in surface sediments more quickly than would happen by natural

sedimentation processes alone. ENR is used in areas with less sediment contamination than those

designated for dredging and capping and only in Recovery Category 2 and 3 areas. In some areas,

ENR may be combined with in situ treatment; i.e., the sand layer may be amended with activated

carbon or other sequestering agents to reduce the bioavailability of organic contaminants such as

PCBs. The effectiveness and potential impacts of using in situ treatment or amendment

technologies, as well as the areas best suited for these technologies, will be evaluated in pilot

studies performed before or during remedial design.

The non-engineered technologies common to all alternatives include: monitored natural recovery,

monitoring, and institutional controls, as described below:

Monitored Natural Recovery (MNR) – Monitored natural recovery relies on natural processes

to reduce ecological and human health risks to acceptable levels, while monitoring recovery of


88

sediments over time to determine remedy success. Within the LDW, natural burial of

contaminants through sedimentation from upstream is the primary natural recovery mechanism.

The sediment transport model (STM) and bed composition model (BCM), supported by RI/FS

data, were used to estimate reduction of sediment COC concentrations over time through natural

recovery.

Refinement of MNR as described in the FS: Terminology used to describe MNR in this ROD

for the Selected Remedy differs from that used in the FS, as follows.

o In the FS the term "MNR" referred only to reduction of COC concentrations through natural

processes until the concentrations reach RAO 3 cleanup levels (benthic SCOs); that is, MNR

would be applied only in areas where concentrations are above the benthic SCO; once benthic

SCOs are reached, MNR would no longer apply and the area would be designated “long-term

monitoring.” As used in the FS, MNR included more intensive monitoring and additional

actions in any areas where benthic SCOs are not achieved within 10 years after remedial

action. Areas where COC concentrations are already below the benthic SCO were designated

in the FS as "long-term monitoring" areas with a lower sampling density, although the FS

acknowledged that reduction of COC concentrations through natural recovery would also

continue in those areas.

o In this ROD (and in LDWG 2012b and LDWG 2013a), the term “MNR” is applied to all

areas where reduction of COC concentrations through natural recovery is predicted to

continue after cleanup is complete (i.e., in areas where concentrations are above benthic

SCOs as well as areas where they are below benthic SCOs). For the Selected Remedy only,

this ROD refines MNR, dividing it into two categories: 1) MNR To Benthic SCO, for areas

where MNR would be used to achieve cleanup levels for RAO 3 (benthic SCO); and 2) MNR

Below Benthic SCO for areas where MNR is used to further reduce COC concentrations to

the remaining cleanup levels for RAOs 1, 2, and 4. MNR To Benthic SCO includes additional

actions (see Section 13.2.2) to be implemented if the SCO is not achieved within 10 years

after remedial action; MNR Below Benthic SCO does not require additional actions in this

ROD if cleanup levels are not achieved. This terminology is more fully described in Section

13.

Monitoring – Monitoring includes sampling sediments, porewater, surface water, and fish and

shellfish tissue to assess site conditions before, during, and after cleanup. All alternatives include

baseline monitoring during the remedial design phase. Monitoring will continue through

construction to assess compliance with construction performance standards, and will continue

over the long term to determine whether technologies are operating as intended and to assess

progress toward achieving the cleanup levels.

Institutional controls – Because none of the alternatives evaluated in the FS would provide

sufficient risk reduction to allow for unrestricted use of the LDW, and to ensure compliance with

the MTCA ARAR requiring institutional controls whenever hazardous substances remain above

cleanup levels (WAC 173-340-440(4)), all the FS alternatives included institutional controls

(ICs). It is important to recognize that even if all natural background-based sediment cleanup

levels and tissue targets were met (keeping in mind that calculated risk-based concentrations are

more stringent than background levels), this would not safely allow for unrestricted use of the


89

LDW (human consumption of unlimited quantities of resident fish and shellfish). The ICs

considered for the LDW include:

o Informational devices, such as seafood consumption advisories, and other ICs intended to

minimize risk to the LDW fishing community, and monitoring and notification of waterway

users, including use of the state’s Environmental Covenants Registry; and

o Proprietary controls, such as environmental covenants, to protect the integrity of the

engineered features such as sediment caps. They would typically require EPA approval prior

to activity that may disturb or encounter contamination that remains in the LDW after

cleanup.

Institutional controls will only be relied upon to the minimum extent practicable, consistent with MTCA

institutional control regulations (WAC 173-340-440(6)).

Governmental controls such as permits required to dredge or fill in the waterway will provide additional

protection.

9.4 Remedial Alternatives

The alternatives use varying combinations of the technologies listed above. Elements that vary among

alternatives include: 1) extent of the active remediation, 2) technologies assigned, and 3) areas where a

technology may be applied as defined by COC concentrations (RALs).

Each of the twelve remedial alternatives is briefly described. Higher numbered alternatives must achieve

progressively lower RALs and they have increasingly larger cleanup footprints (e.g., the cleanup footprint

for Alternative 3 is larger than that for Alternative 2).

For the alternatives that emphasize removal (the "R" alternatives), dredging/excavation and disposal

would be the primary technologies used for active remediation. The combined technology ("C")

alternatives emphasize the use of capping, enhanced natural recovery (ENR), and in situ treatment. The C

alternatives would use dredging and excavation only where capping and ENR/in situ treatment are not

feasible due to requirements to maintain water depths in habitat areas, the navigation channel, or berthing

areas. In the C alternatives, ENR is used only in areas with low scour potential and moderate sediment

contaminant concentrations because underlying sediment contamination is not isolated by this technology

as it is with a cap. For the FS and this ROD, moderate contamination is defined as 1 to 1.5 times the

intertidal RALs (applied over the top 45 cm) and 1 to 3 times the subtidal LDW-wide RALs (applied over

the top 10 cm). More aggressive technologies such as caps would be used in highly contaminated areas

(where concentrations are greater than 1.5 times the intertidal RAL or 3 times the LDW-wide RAL) and

in areas with scour potential. Dredging, and partial dredging and capping, would be used where water

depth constraints preclude capping alone.

The areas addressed by the cleanup alternatives as depicted in the FS and this ROD (Figure 13) are

preliminary. The sediment contaminant concentrations used to delineate the areas addressed in FS

remedial alternatives were collected over a 20-year period, from 1991 to 2010, with the bulk of the data

collected prior to 2005. Sampling to be conducted during remedial design may establish different

sediment contaminant concentrations; for example, some areas may have already recovered naturally

within that time while others may have become more contaminated due to ongoing input from


90

contaminant sources. The specific areas to be addressed by remedial technologies and MNR will be

refined based on results from additional sampling during remedial design.

Because all alternatives use similar technologies, the primary ARARs are the same for all alternatives.

ARARs compliance is described in Sections 10.1.2 (for FS alternatives generally) and 14.2 (for the

Selected Remedy). All Alternatives, including the Selected Remedy (except Alternative 1, No Action),

include off-site disposal of dredged material. Data from the RI/FS indicate that sediment removed from

the LDW can be disposed of in a solid waste landfill that is compliant with RCRA Subtitle D. If wastes

that require disposal in a landfill permitted to receive RCRA hazardous wastes or Toxic Substances

Control Act (TSCA) regulated wastes are encountered during remedial design or remedial action, they

will be disposed in a landfill compliant with RCRA Subtitle C or TSCA. Alternative 2 used a different

disposal technology, contained aquatic disposal, which would have made Section 404 of the Clean Water

Act a more important ARAR if that alternative had been selected. Only Alternative 5R-Treatment used

soil washing; however, ARARs for disposal or beneficial reuse of treated material would be the same as

for disposal of untreated sediments.

Table 22 (above) summarizes the RALs and associated types of actively-remediated areas under each

alternative, along with the total area of active remediation for each. Table 24 summarizes the areas and

volumes associated with each remedial technology for each of the alternatives as well as costs and

construction durations. In the FS, net present values costs for cleanup alternatives were calculated at a

2.3% discount rate based upon guidance from the Office of Management and Budget (OMB).17,18

Table

25 provides a comparison of costs calculated at discount rates of 0%, 2.3%, and 7%. The cost of

implementing in-waterway cleanups at the EAAs is estimated at $95-100 million; this cost is not included

in the cost estimates for the alternatives.

For each alternative, Figure 14 shows the construction period and the projected time to achieve a range of

risk reduction benchmarks. Figure 14 shows time to achieve RAOs for RAOs 3 and 4, and time to achieve

RAOs for all COCs except arsenic for RAO 2. For RAO 1 (all COCs) and RAO 2 (arsenic), as discussed

in Section 8.2.1, it was not possible to project the time to achieve cleanup levels using RI/FS models.

Instead, Figure 14 shows the time to reach the lowest model-projected concentrations. For informational

purposes Figure 14 also shows other information, such as time to achieve 1 x 10-5

excess cancer risk for

RAOs 1 and 2, to show when progress would be made towards achieving these RAOs.

17

Appendix C of Office of Management and Budget Circular A-94 for the Year 2011. 18

EPA's Guide to Developing and Documenting Cost Estimates during the Feasibility Study, (EPA 2000a)

recommends that a discount rate of 7% be used for estimating the net present value of cleanups conducted by

non-federal parties. The rate of 7% approximates the marginal pretax rate of return on an average investment in

the private sector adjusted for expected inflation. The lower discount rate was included because EPA believes a

rate derived from interest rates published in Appendix C of OMB Circular A-94 better reflects current economic

conditions for safely setting aside money for future cleanup costs. The lower discount rate was also incuded to

reflect differing costs of capital for public entities, because, as stated by LDWG in the FS, local governments are

likely to be involved in the implementation of the remedy (LDWG 2012a).


91

Table 24. Remedial Alternative Areas and Volumes

Site -wide Remedial

Alternative

Remedial Alternative Technology and Areas

Total Dredge

Volume (cy)

Construction

Period

(years)

Net Present

Value Cost

($Millions)

EAAs

(acres)a

Dredge

(acres)

Partial

Dredge and

Cap (acres)

Cap

(acres)

ENR/

in situ

(acres)

MNR To

Benthic

SCOb

(acres)

MNR (in Alt 5C

Plus, MNR

Below Benthic

SCO) (acres)

Total

Active

Remedy

(acres)

1 No Further Action 29 0 0 0 0 0 412 0 n/a n/a $9

2 Removal 29 29 3 0 0 148 232 32 580,000 4 $220

2 Removal with CAD 29 29 3 0 0 148 232 32 580,000 4 $200

3 Removal 29 50 8 0 0 122 232 58 760,000 6 $270

3 Combined Technology 29 29 8 11 10 122 232 58 490,000 3 $200

4 Removal 29 93 14 0 0 73 232 107 1,200,000 11 $360


5 Removal 29 143 14 0 0 23 232 157 1,600,000 17 $470

5 Removal with Treatment 29 143 14 0 0 23 232 157 1,600,000 17 $510


Alternative 5 Combined Technology Plus

29 64 20 24 48 33 223 156 790,000 7 $305

Selected Remedyc 29 85 20 24 48 33 202 177 960,000 7 $342

6 Removal 29 274 28 0 0 0 110 302 3,900,000 42 $810

6 Combined Technology 29 108 42 51 101 0 110 302 1,600,000 16 $530

a. The 29 acres addressed by the EAAs are not included in area estimates for remedial alternatives. b. Includes areas that the FS predicted will have naturally recovered enough that concentration levels are below the benthic SCO by the time sampling is conducted for remedial design (called

"verification monitoring" in the FS). c. Selected Remedy includes changes from the Proposed Plan in dredge areas, volumes, and overall costs based on public comment and other considerations, see Section 12 for details.


92

Table 25. Summary of Remedial Alternative Costs ($Millions)

Discount Rate Cost Parameter

Remedial Alternative

1 2R

2R-

CAD 3R 3C 4R 4C 5R 5R-T 5C

5C

Plus

Selected

Remedy 6R 6C

0% (No discount)

Capital Costs NA — — — — — — — — — $280 $327 — —

OM&M, Reporting, Agency

Oversight Costs NA — — — — — — — — — $68 $68 — —

Total Cost (2011 dollars) NA $250 $230 $310 $230 $430 $300 $580 $630 $330 $348 $395 $1,300 $650

2.3% (30-yr OMB 2011 Real Interest

Treasury Rate)

Capital Costs NA $169 $148 $224 $156 $324 $221 $430 $473 $250 $258 $295 $771 $478


Oversight Costs NA $46 $49 $43 $46 $39 $41 $37 $37 $41 $48 $48 $42 $51

Total Net Present Value

Cost NA $220 $200 $270 $200 $360 $260 $470 $510 $290 $305 $342 $810 $530

7% (per EPA's (2000) Guide to

Developing and Documenting Cost

Estimates during the Feasibility

Study)

Capital Costs NA — — — — — — — — — $220 $243 — —


Oversight Costs NA — — — — — — — — — $28 $28 — --

Total Net Present Value Cost NA $180 $160 $210 $170 $270 $220 $320 $350 $240 $248 $270 $410 $370

Notes:

Costs are reported in $ million. Total costs are rounded to 2 significant digits.

Costs for all alternatives do not include costs to conduct cleanup in the EAAs.

In Appendix I of the FS, monitoring, operations and maintenance, reporting, and agency oversight activities were assumed to occur for a period of 30 years from the start of construction for all of the alternatives except Alternative 6R, for which those activities were assumed to occur for a period of 45 years from the start of construction. In reality, it is not known how long these activities would continue. If they were to occur for more than 30 years (or 45 years in the case of Alternative 6R), the associated costs would be higher than these estimates, although on a net present value basis they may not add appreciably to these estimated costs because they are so far in the future.

See Table 29 for a detailed cost summary for the Selected Remedy.


93

Figure 14. Time to Achieve Risk Reduction for All Alternatives


94



95

Alternative 1 – No Further Action – This alternative would not implement any further action following

removal or capping implemented through Early Actions, with the exception of continued LDW-wide

monitoring. It includes no ICs other than the existing seafood consumption advisories and those

implemented for the Early Actions. This alternative provides a baseline to compare the other remedial

alternatives against; its inclusion is required by CERCLA. LDW-wide monitoring costs at net present

value (NPV) (using a 2.3% discount rate) are estimated to be $9 million for Alternative 1.

Alternatives 2R and 2R-CAD – These alternatives would

actively remediate 32 acres with contaminant

concentrations above the Alternative 2 RALs (Table 22).

The area and volume of contaminated sediments remediated

by each technology and estimated costs are provided in

Table 24. Figure 14 shows the time projected by FS models

to achieve risk benchmarks for each remedial action

objective. Alternatives 2 and 2R-CAD include:

For areas with COC concentrations exceeding the

RALs, Alternative 2R includes dredging with

upland landfill disposal, while Alternative 2R-CAD

adds contained aquatic disposal (CAD) to address

disposal of some of the dredged material.

In areas with COC concentrations below the RALs,

MNR would be used to reduce COC concentrations

to the RAO 3 cleanup levels (benthic SCO criteria)

as well as to achieve FS cleanup objectives19

for

RAOs 2 and 4. As noted above, the FS made no distinction between MNR To Benthic SCO and

MNR Below Benthic SCO, so these terms are not used for any alternative except the Selected

Remedy (5C Plus as modified for this ROD). As discussed in Section 9.3, the FS used the term

"MNR" to include enhanced monitoring and additional actions only for any area where COC

concentrations are not reduced to the benthic SCO levels. For simplicity, this ROD uses the term

MNR to refer to all areas where reduction in COC concentrations is predicted, regardless of

whether levels are above or below the benthic SCO.

For all alternatives, effective and appropriate ICs will be used to reduce exposure of fish

consumers to contamination in fish and shellfish. Examples include seafood consumption

advisories, outreach, and education programs.

19. The FS was written before promulgation of the revised SMS, which created a new term “sediment cleanup

objective” (SCO). The term "cleanup objectives" was used in the FS to mean the PRG or as close as practicable

to the PRG where RI/FS models could not predict acheivement of a PRG. The FS used long-term model-

predicted concentrations as estimates of “as close as practicable” to PRGs. This ROD uses the term “FS cleanup

objective” when referring to the terminology in the FS and “sediment cleanup objective” or SCO when referring

to the terminology in the revised SMS.

Remedial Action Objectives

RAO 1: Reduce risks associated with the

consumption of contaminated resident LDW fish

and shellfish by adults and children with the

highest potential exposure to protect human

health.

RAO 2: Reduce risks from direct contact (skin

contact and incidental ingestion) of contaminated

sediments during netfishing, clamming, and beach

play to protect human health.

RAO 3: Reduce to protective levels risks to

benthic invertebrates from exposure to

contaminated sediments.

RAO 4: Reduce to protective levels risks to crabs,

fish, birds, and mammals from exposure to

contaminated sediment, surface water, and prey.


96

The 2R/2RCAD alternatives were designed to achieve the following at a minimum, relative to the RAOs

for the In-waterway Portion of the Site:

For RAO 1 (human health seafood consumption): Incremental risk reduction through active

remediation and further risk reduction through MNR. RI/FS models predict it will take 24 years

after the start of construction to reach lowest model-predicted values (FS cleanup objectives).

For RAO 2 (human health direct contact): Meet FS cleanup objectives within 10 years following

construction completion.

For RAO 3 (protection of benthic community): Reduce contaminants in sediment to meet the

benthic CSL within 10 years following construction, and the benthic SCO within 20 years

following construction. (See Section 5.3.1.1 for additional information on the benthic CSL and

benthic SCO).

For RAO 4 (protection of river otter): Meet cleanup level within 10 years following construction.

Alternatives 3R and 3C – These Alternatives actively remediate 58 acres with contaminant

concentrations above the Alternative 3 RALs (Table 22). The area and volume of contaminated sediments

remediated by each technology and estimated costs are provided in Table 24. A greater amount of surface

and subsurface contamination is removed by these alternatives than by Alternative 2R/2R-CAD, and they

rely more upon active remediation to reduce risks to human health from consuming contaminated seafood

than the previous alternatives. Alternatives 3R and 3C include:

For areas exceeding the RALs, Alternative 3R has a removal emphasis (i.e., dredging) with

upland disposal/MNR, and Alternative 3C uses a combined technology approach (i.e., capping

and ENR/MNR/in situ treatment) in addition to dredging with upland disposal.

MNR is used in areas with concentrations below RALs to achieve the benthic SCO within 20

years following construction, with additional COC concentration reduction over time to the FS

cleanup objectives.

ICs would be used as described in Alternative 2R/2R-CAD.

These alternatives are designed to achieve, at a minimum, the outcomes of Alternative 2R/2R-CAD, plus:

For RAO 1: Achieve greater permanence through actively remediating a larger area, and relying

less on natural recovery. RI/FS models predict it will take 18 years (3C) or 21 years (3R) years


For RAOs 2 and 4: Achieve FS cleanup objectives immediately following construction, rather

than 10 years following construction.

For RAO 3: Achieve the benthic CSL immediately following construction, rather than 10 years

following construction. The benthic SCO would still not be projected to be reached for 20 years.

Alternatives 4R and 4C – These alternatives actively remediate 107 acres with contaminant


remediated by each technology and estimated costs are provided in Table 24. MNR is used in areas with

concentrations below the RALs to achieve the benthic SCO within 10 years following construction, with

additional COC concentration reduction over time to the FS cleanup objectives. ICs would be used as


97

described in Alternative 2. Alternatives 4C and 4R rely more on active remediation than lower numbered

alternatives to reduce COC concentrations. These alternatives are designed to achieve, at a minimum, the

outcomes of Alternative 3, plus:

For RAO 1: Achieve greater permanence through actively remediating a larger area than lower

numbered alternatives, and relying less on natural recovery. RI/FS models predict it will take 21

years (for 4C and 4R) after the start of construction to reach lowest model-predicted values (FS

cleanup objectives).

For RAOs 2 and 4: Same as Alternative 3R/3C.

For RAO 3: Achieve the benthic SCO for within 10 years following construction as opposed to

20 years following construction.

Alternatives 5R, 5R-Treatment, and 5C – These alternatives actively remediate 157 acres with

contaminant concentrations above the Alternative 5 RALs (Table 22). The area and volume of

contaminated sediments remediated by each technology and estimated costs are provided in Table 24.

These three alternatives do not use MNR to reach the benthic SCO20

; however, natural recovery is relied

upon to further reduce risks to the FS cleanup objectives. Alternative 5R-Treatment utilizes removal with

ex situ treatment (soil washing) and disposal/re-use. These three alternatives rely more on active

remediation than lower numbered alternatives to reduce COC concentrations. These alternatives are

designed to achieve, at a minimum, the outcomes of Alternative 4 R/4C, plus:

For RAO 1: Achieve greater permanence by actively remediating a larger area than lower-

numbered alternatives. RI/FS models predict it will take 22 years (5R/5R-T) or 17 years (5C)



For RAO 3: Achieve the benthic SCO immediately following construction as opposed to 10 years

following construction.

Alternatives 6R and 6C – These alternatives actively remediate 302 acres with contaminant


remediated by each technology and estimated costs are provided in Table 24. Alternative 6R has a

dredging emphasis with upland disposal, while Alternative 6C emphasizes combined technologies

including ENR/capping where appropriate, in addition to dredging with upland disposal. These

alternatives are designed to achieve, at a minimum:

For RAO 1: Achieve the lowest model-projected COC concentrations immediately after

construction is completed, rather than relying on MNR. The FS estimates it will take 42 years

(6R) or 16 years (6C) after the start of construction to reach lowest model-predicted values (FS

cleanup objectives).


20. Although Table 24 shows 23 acres of MNR, the FS predicts that COC concentrations in these areas will be

reduced to the SCO prior to the start of construction.


98

For RAO 3: Achieve the benthic SCO immediately following construction as opposed to 10 years

following construction.

Alternatives 6R and 6C would rely the most on active remediation to reduce COC concentrations relative

to all other alternatives.

Alternative 5C Plus – Alternative 5C Plus in the Proposed Plan was developed by modifying FS

Alternative 5C to include additional remedial elements as described in LDWG (2012b), which evaluated

these additions to address several concerns, including the need for:

Additional RALs for subsurface sediments in areas outside of Recovery Category 1 areas to

address the potential that subsurface contamination could be disturbed and exposed at the surface

through activities such as emergency or high-power vessel operations, vessel groundings,

maintenance activities, or earthquakes. These areas are described in more detail in Table 22

footnote a, and in Section 13.

Additional dredging in shoaled areas of the navigation channel where COC concentrations exceed

RALs in the top 2 ft to address the potential that subsurface contamination could be disturbed

through maintenance dredging.

Increased cap thickness to 4 ft in intertidal clamming areas to provide adequate habitat for clams.

Increased sediment monitoring to evaluate natural recovery progress in areas where COC

concentrations are below the SCO but above cleanup levels (designated as MNR Below SCO in

this ROD; see Section 13.2.2).

In addition, LDWG (2012b) evaluated greater use of MNR to reduce concentrations of non-human health

COCs in surface sediments, while continuing to use active remediation when RALs for human health

COCs are exceeded. Six scenarios were developed by LDWG (2012b). EPA, in coordination with

Ecology, selected Scenario 5a (referred to as Alternative 5C Plus) in the Proposed Plan.

Estimates of cleanup areas and volumes for Alternative 5C Plus were modified from those in Alternative

5C in LDWG (2013a). Alternative 5C Plus actively remediates 156 acres with contaminant concentrations

above the Alternative 5C Plus RALs (Table 22). These RALs are the same as for Alternative 5C, except

as follows.

New RALs for subsurface PCBs were added for potential scour areas within Recovery Category 2

and 3 areas (see Section 13). The new PCB RALs address the concern that high concentrations of

PCBs, the most prevalent COC in the LDW, could become exposed through human activities

such as digging in the beach in intertidal areas or emergency ship maneuvering in subtidal areas.

The PCB RAL, for subsurface sediments in intertidal areas, is 65 mg/kg OC (e.g., 1.3 mg/kg

PCBs at 2% TOC) applied as an average over the top 45 cm, and the PCB RAL for subsurface

sediments in subtidal areas is 195 mg/kg OC (e.g., 3.9 mg/kg PCBs at 2% TOC) applied as an

average over the top 10 cm. No other alternatives have subsurface RALs for PCBs within

Recovery Category 2 and 3 areas.

The RALs for non-human health COCs in surface sediments were increased to 2 times the

benthic SCO, not to exceed the CSL, in Recovery Category 2 and 3 areas. The benthic SCO must

be met within 10 years of completing remedial action.


99

The estimated area and volume of contaminated sediments remediated by each remedial technology and

the estimated costs are provided in Table 24. Alternative 5C Plus includes 33 acres of MNR To Benthic

SCO, and 223 acres of MNR Below Benthic SCO (with more monitoring than in the FS Alternatives) for

RAO 1. Alternative 5C Plus would rely more on active remediation than 5C (but less than 6C) to reduce

COC concentrations in surface sediments. Alternative 5C Plus is designed to achieve, at a minimum, the

following outcomes:

For RAOs 1, 2 and 4: Achieve greater risk reduction than 5C because there would be a larger

volume of sediments actively remediated, and an increased emphasis on reducing high

concentrations of PCBs in subsurface sediments.

For RAO 3: Achieve the benthic CSL immediately following construction, and the benthic SCO

within 10 years following construction.

Selected Remedy - In the Selected Remedy, additional changes were made to Alternative 5C Plus as it

was presented in the Proposed Plan to address public comments. These changes are summarized here and

discussed in more detail in Sections 12 and 13.

Estimates of dredging areas and volumes were revised using new data collected after completion

of the FS.

RALs for non-human health COCs were modified to 2 times the benthic SCO (and "not to exceed

CSL" was omitted).

The required space between the authorized navigation channel depth and the top of caps in the

federal navigation channel (sometimes called the “buffer”), was changed from 3 ft to 4 ft.

Additional dredging in shoaled areas of the navigation channel where COC concentrations exceed

RALs at any depth that is not at least 2 ft deeper than the authorized channel depth (this allows

for over-dredging associated with channel maintenance) This additional dredging addresses the

potential that subsurface contamination could be disturbed through maintenance dredging.

Revised area, volume, and cost estimates for the Selected Remedy are shown in in Tables 24 and 25.

Although these changes may affect the time to complete the remedial action, EPA retained the 5C Plus

estimate of 7 years, because these additional volume estimates are uncertain due to the limited number of

cores used to characterize these additional areas, which would need to be addressed during remedial

design. Additionally, the time to complete source control activities, length of the fish window and number

of dredges operating at the same time may affect construction duration. The Selected Remedy remains

sufficiently similar to 5C Plus that the analysis for 5C Plus (Section 10) is also applicable to the Selected

Remedy.


100

10 Summary of Comparative Analysis of Alternatives

EPA used the nine criteria required by CERCLA and the NCP to evaluate and select a remedy for the In-

waterway Portion of the LDW Superfund Site. This section describes the relative performance of each

alternative against the nine criteria, noting how the Selected Remedy, 5C Plus, compares to the other

alternatives. The nine criteria are in three categories: threshold criteria, primary balancing criteria, and

modifying criteria. The findings and recommendations in EPA’s EJ Analysis (EPA 2013a) were

considered as part of the nine criteria analysis, as discussed in Section 10.3.3.

Nine Criteria for CERCLA Remedy Selection

Threshold criteria. Each alternative must meet threshold criteria to be eligible for selection.

Overall Protection of Human Health and the Environment — addresses whether each alternative provides

adequate protection of human health and the environment and describes how risks posed through each

exposure pathway are eliminated, reduced, or controlled through treatment, engineering controls, and/or

institutional controls.

Compliance with Applicable or Relevant and Appropriate Requirements (ARARs) — CERCLA Section

121(d) and NCP §300.430(f)(1)(ii)(B) require that remedial actions at CERCLA sites at least attain legally

applicable or relevant and appropriate federal and state requirements, standards, criteria, and limitations which

are collectively referred to as ARARs, unless such ARARs are waived under CERCLA Section 121(d)(4).

Primary balancing criteria. Balancing criteria are used to evaluate the major technical, cost, and other trade-offs

among the various remedial alternatives.

Long-Term Effectiveness and Permanence — refers to expected residual risk and the ability of a remedy to

maintain reliable protection of human health and the environment over time, once cleanup levels have been

met. This criterion includes the consideration of residual risk that will remain onsite following remediation and

the adequacy and reliability of controls.

Reduction of Toxicity, Mobility, or Volume Through Treatment — refers to the anticipated performance of

the treatment technologies that may be included as part of a remedy.

Short-Term Effectiveness — addresses the period of time needed to implement the remedy and any adverse

impacts to workers, the community, and the environment during construction and operation of the remedy until

cleanup levels are achieved (and how they may be mitigated).

Implementability — addresses the technical and administrative feasibility of a remedy from design through

construction and operation. Factors such as availability of services and materials, administrative feasibility,

and coordination with other governmental entities are also considered.

Cost — addresses the cost of construction and any long term costs to operate and maintain the alternative, in

terms of estimated capital, annual operation and maintenance, and total net present value costs.

Modifying criteria. Modifying criteria are based on the public comments received on the Proposed Plan,

discussions with the state, and consultation with affected Tribes.

State /Tribal acceptance — Assessment of State concerns including (1) The state’s position and key

concerns related to the Selected Remedy and other alternatives; and (2) State comments on ARARs or the

proposed use of ARAR waivers. Concerns of affected Tribes are also considered.

Community acceptance — This assessment includes determining which components of the alternatives

interested persons in the community support, have reservations about, or oppose.


101

10.1 Threshold Criteria

An alternative must meet both threshold criteria to be eligible for selection as a remedial action.

10.1.1 Overall Protection of Human Health and the Environment

All of the alternatives provide a substantial reduction in risk when compared to baseline conditions. They

meet the threshold criterion of overall protection of human health and the environment over varying

timeframes. The objective of Alternatives 2 - 6 is to reach their differing RALs by the end of construction,

and cleanup levels in the long term. Alternatives vary in: 1) the extent of reliance on institutional controls

to reduce exposure, especially from consumption of resident LDW seafood, 2) the model-projected time

to achieve short- and long-term risk reduction (RI/FS models predicted that Alternatives 2-6 would reach

the lowest model-predicted concentrations and associated risks in 16 to 42 years, as shown in Figure 14),

and 3) the extent to which they depend on potentially less reliable or permanent technologies such as

ENR and natural recovery to achieve risk reduction. Alternatives that depend more on ENR and MNR

instead of dredging and capping have a greater risk of subsurface contamination becoming exposed and

increasing long-term COC concentrations, as discussed in Section 10.2.1. All of these factors provide

more certainty that alternatives that rely more on dredging and capping (e.g., Alternatives 5 and 6) will

achieve long-term risk reduction than those that rely more on natural recovery (e.g., Alternatives 1 – 3).

As discussed in Section 8.2.1, the RI/FS models (STM, BCM, and food-web model) could not predict

that, for any alternative, LDW sediment and tissue COC concentrations would reach the risk- and natural

background-based sediment cleanup levels and target tissue concentrations. Although RI/FS data

indicated strong evidence for natural recovery in some areas of the waterway, our ability to predict the

rate and risk reduction achieved by natural recovery processes 30 – 40 years in the future is limited. EPA

is therefore selecting the cleanup levels shown in Table 19 and Table 20 as the long-term objectives for

the cleanup.

EPA’s intent is for the Selected Remedy to achieve risk reduction and protectiveness while minimizing

reliance on seafood consumption-related Institutional Controls to the extent practicable. Alternative 1

would not protect human health and the environment. It does not include any active cleanup or

institutional controls beyond the current Washington Department of Health (WDOH) health advisory, and

those controls implemented at EAAs. Therefore, it is not discussed further.

10.1.1.1 Protecting seafood consumers (RAO 1)

Risks associated with PCBs. Figure 15 shows estimated cancer and noncancer risks for a variety of

seafood consumption rates for PCBs at baseline, as predicted by RI/FS models, and at the target tissue

concentrations. Only PCBs could be addressed in the RI/FS food-web model because RI data did not

provide sufficient information to develop predictive relationships between sediment concentrations and

tissue concentrations for other human health COCs; in particular, arsenic and cPAHs did not show close

relationships, and there were insufficient tissue data for dioxins/furans to do so. Calculating health risks

from eating fish or shellfish is critically dependent on how much and what type of seafood a person may

eat. PCB risks for adult Tribal consumers of seafood at RME consumption rates (approximately 3 meals

per week) are estimated to be 5 x 10-6

and HQ of less than 1 if target tissue concentrations are achieved.

The FS estimated an adult Tribal RME excess cancer risk of 2 x 10-4

and noncancer HQ of 4 - 5, and a

child tribal RME excess cancer risk of 3 x 10-5

and a noncancer HQ of 9 - 10 for PCBs only at the model-


102

predicted steady state for all alternatives. These estimates represent a post-cleanup reduction in PCB risks

of approximately 90% at the model-predicted steady state and 99% at the target tissue concentration for

the adult Tribal RME seafood consumption rate when compared to baseline risks. As shown in Figure 15,

PCB-only risks would be lower for those who consume less seafood.

Risks associated with all COCs. Figure 16 shows estimated total cancer and noncancer risks at the target

tissue concentrations for PCBs at a range of consumption rates for different seafood types. None of the

alternatives would meet MTCA risk thresholds (Section 8.2.1) for consumption of resident fish and

shellfish at the consumption rates reported for Tribal or Asian Pacific Islander populations, even if they

were to meet the target tissue concentrations, because target concentrations for some COCs are based on

natural background levels that are higher than calculated protective risk-based levels (RBTCs-see Table

21). Cancer risks at the adult tribal RME seafood consumption rate (approximately 3 meals per week) if

fish and shellfish target tissue concentrations are achieved for all COCs are estimated to be 3 x 10-4

,

which is above the excess cancer risk thresholds for both CERCLA and MTCA. Cancer risks at the child

Tribal RME seafood consumption rate are estimated to be 5 x 10-5

. The main contaminants contributing to

these post-cleanup risks are arsenic and dioxins/furans. The HQ for noncancer risks at both the adult and

child tribal RME seafood consumption rate would be less than the CERCLA and MTCA threshold of 1

(based on the risks associated with PCBs, the COC with the highest HQ for adults, and with arsenic, the

COC with the highest HQ for children).

At a consumption rate of one meal every other month, seafood consumption risks for all seafood types

except crab whole body are estimated to be at or below the MTCA excess cancer risk multiple

contaminant threshold of 1 x 10-5

and HI of 1 at target tissue concentrations (if the hepatopancreas were

removed, crab edible meat would also be below risk thresholds). At a consumption rate of one meal per

month, excess cancer risks for all food types are estimated to be below the CERCLA excess cancer risk

threshold of 1 x 10-4

and noncancer HI of 1.

10.1.1.2 Reducing Direct Contact Risks (RAO 2)

For direct contact with sediments in netfishing, clamming, and beach play areas (RAO 2), all alternatives

are predicted to result in risks that are within the CERCLA risk range and meet the minimum MTCA risk-

reduction requirements: 1) a total excess cancer risk of less than 1 x 10-5

cumulatively for all COCs; 2)

excess cancer risks for individual COCs less than or equal to 1 x 10-6

(except for arsenic), and 3)

noncancer HI less than or equal to 1. The natural recovery model predicts arsenic will reach an excess

cancer risk range below 1 x 10-5

but above 1 x 10-6

. Alternative 2 requires a period of natural recovery to

meet these objectives, whereas Alternatives 3 – 6 meet them immediately after construction, and thus

there is greater likelihood of a permanent reduction in contamination and risk.


103

Exposure assumptions: Excess cancer risks and noncancer hazard quotients (HQs) were calculated assuming market basket consumption using the exposure assumptions for the adult tribal RME seafood consumption scenario. Three meals per week is approximately equal to the consumption rate used for the adult Tribal RME Scenario in the HHRA. A meal is equal to 8 ounces.

Excess cancer risks and noncancer HQs shown in this figure are only for total PCBs; the calculation of total risks for the site would include all contaminants. Excess cancer risks and noncancer HQs were calculated using a) LDW baseline tissue concentrations from the LDW HHRA; b) model-predicted steady-state tissue concentrations, based on predictions of tissue concentrations using the calibrated LDW food web model at a sediment concentration of 40 µg/kg dw and a water concentration of 0.6 ng/L; and c) target tissue levels, based upon either the higher of either non-urban Puget Sound tissue concentrations or the species-specific tissue RBTCs.

Risk thresholds: In the top portion of the figure showing the excess cancer risks, the shaded area indicates EPA's acceptable excess cancer risk range under CERCLA. The dotted line indicates the MTCA threshold for individual carcinogens. In the lower portion of the figure showing noncancer HQs, the dotted line indicates the CERCLA and MTCA noncancer threshold.

Figure 15. Comparison of Total PCB Excess Cancer Risks and Noncancer HQs for Seafood Consumption Calculated using LDW Baseline, Model-predicted, and Target Tissue Concentrations


104

Excess cancer risks and noncancer hazard quotients (HQs) were calculated using the exposure assumptions for the adult tribal RME seafood consumption scenario. One meal is equal to 8 ounces, and 3 meals per week is approximately equal to the rate used for the adult tribal RME scenario in the HHRA.

Target tissue concentrations are the higher of either non-urban Puget Sound tissue concentrations or species-specific risk-based threshold concentrations

Excess cancer risks were calculated as the sum of excess cancer risks for inorganic arsenic, cPAHs, total PCBs, and dioxins/furans, depending on whether target concentrations were available for a given contaminant-species combination. For English sole, perch, and crabs, only PCB and dioxin/furan data were available, as noted by the ǂ in the legend. For calculating market basket consumption, the risks for cPAHs and inorganic arsenic are based only on the consumption of clams because clams account for over 95% of the risk. It should also be noted that a target tissue concentration was not available for dioxins/furans for benthic fish fillet and pelagic fish. Thus, the whole body benthic fish target for dioxins/furans was used as a surrogate for risk calculations.

Noncancer HQs for total PCBs are presented because HQs were the highest for PCBs.

Figure 16. Excess Cancer Risks and Noncancer HQs for Seafood Consumption Calculated Using Target Tissue Concentrations


105

10.1.1.3 Protecting Benthic Communities (RAO 3)

All alternatives are predicted to achieve the RAO 3 cleanup levels for protection of benthic invertebrates

(benthic SCO criteria), within 6 to 20 years. Alternatives 2 to 4 and the Selected Remedy (5C Plus) rely

on MNR to reduce COC concentrations to the benthic SCO, with more reliance on MNR than in the

higher-numbered alternatives and hence a greater likelihood of permanence of contaminant reduction.

The FS assumed additional actions would be undertaken if the benthic SCO was not achieved within 10

years of completion of the remedial action. The Selected Remedy differs from Alternatives 2 – 4 in that it

relies on MNR with contingency actions as discussed above to reduce COC concentrations only for non-

human health COCs.

10.1.1.4 Protecting Wildlife (RAO 4)

All alternatives are predicted to achieve the RAO 4 cleanup levels for protection of wildlife (river otters)

shortly following construction. Alternatives 2 and 3 are predicted to require a short period of natural

recovery to achieve the cleanup levels.

10.1.2 Compliance with ARARs

ARARs for the cleanup alternatives are shown in Table 26. The most significant ARARs for in-waterway

remedial action are the MTCA/SMS requirements and federal and state water quality criteria and

standards, as discussed in Section 8.2.2. As discussed in Section 10.1.1, the RI/FS models indicate that for

all alternatives, the long-term COC sediment concentrations achievable in the In-waterway Portion of the

Site will be limited by the extent to which all ongoing sources, including COCs entering the waterway

from the upstream Green/Duwamish River system and remaining lateral sources, can be controlled in this

urban environment. Also for all alternatives, the ability to meet surface water quality ARARs will be

limited to the extent that all sources, including upstream water quality, can be improved.

Section 14.2 discusses compliance of the Selected Remedy with ARARs. That discussion would apply to

all the FS alternatives with respect to MTCA/SMS, surface water quality, and ESA ARARs compliance.

See the Project Specific Comments column in Table 26 for discussion of compliance with all other

ARARs.

The ARARs selected in this ROD are remedial action requirements for the Selected Remedy. This ROD,

like all other CERCLA RODs, is not a decision document for any other purpose and does not establish

requirements for implementation of source control by other regulatory agencies.


106

[This blank page is intentionally left blank for double-sided printing.]


107

Table 26. Applicable or Relevant and Appropriate Requirements, LDW Superfund Site

Topic Standard or Requirement

Regulatory Citation

Project-Specific Comments Federal State

Hazardous Substance

Cleanup; Sediment

Quality

Washington State cleanup standards;

Marine Sediment Cleanup Standards;

Sediment Cleanup Objectives (SCO);

Cleanup Screening Levels (CSL)

Model Toxics Control Act (MTCA) (RCW

70.105D; WAC 173-340); MTCA Sediment

Management Standards (SMS) (RCW

70.105D; WAC 173-204)

Substantive MTCA requirements that are more stringent than CERCLA requirements are ARARs. A combination of sediment

dredging, capping, enhanced natural recovery (ENR), monitored natural recovery (MNR), and potentially in-situ amendment as

treatment, along with the minimally necessary use of fish and shellfish consumption advisories as ICs to reduce fish and

shellfish consumption, will be employed to meet the substantive requirements of SCO compliance for the protection of human

health, marine benthic invertebrates and higher trophic level species, as set forth in WAC 173-204-560-562, 564 to the extent

technically possible, or without a net adverse environmental impact, and at a minimum, the substantive requirements of CSL

compliance. Institutional Controls (ICs) will be required as set forth in WAC 173-340-440(4)(a).

Surface Water Quality Surface water quality standards. Federal

recommended Ambient Water Quality

Criteria (AWQC); National Toxics Rule

(NTR); State Water Quality Standards

(WQS)

AWQC per Clean Water Act Section

304(a) (33 U.S.C. § 1314(a)) at http://

water.epa.gov/scitech/swguidance/

standards/criteria/current/index.cfm; NTR at

40 CFR 131.36(b)(1) as applied to

Washington, 40 CFR 131.36(d)(14)

Water Pollution Control Act (RCW 90.48);

WQS (WAC 173-201A); Aquatic Life

Criteria (ALC) numerical criteria (WAC

173-201A-240)

Sediment remediation described immediately above will improve surface water quality to an unknown degree in combination

with source control implementation under state-lead authority. Surface water concentrations shall be at least as stringent as all

of the following: 1) all WQS in WAC 173-201A; 2) AWQC unless it can be demonstrated that such criteria are not relevant and

appropriate for the LDW or for a specific hazardous substance; and 3) the NTR. See WAC 173-340-730(3)(b), consistent with

Sections 121(d)(2)(A)(ii) and (B)(i) of CERCLA and 40 CFR 300.430(e).

Solid Waste Disposal Requirements for solid waste handling

management and disposal

Solid Waste Disposal Act (42 U.S.C. 6901-

6992K; 40 CFR 257-258)

Solid Waste Management

(RCW 70.95; WAC 173-350)

Substantive requirements for non-dangerous or non-hazardous waste generated during remedial activities unless wastes meet

recycling or other exemptions will be complied with.

Waste Treatment,

Storage, and Disposal

Dangerous or Hazardous Waste

Management

Resource Conservation and Recovery Act,

Hazardous Waste (42 U.S.C. §§ 6901-

6992K, 40 CFR 260-279)

Dangerous Waste Management (RCW

70.105; WAC 173-303)

Dredged materials contains solid waste subject to solid waste handling requirements above. It would also be

hazardous/dangerous waste if it contained a listed waste or displayed a hazardous waste characteristic (e.g., per Toxicity

Characteristic Leaching Procedure). Based on the Remedial Investigation (RI), hazardous/dangerous waste is not anticipated in

LDW sediments. If it is encountered 40 CFR Part 262 generator rules in Washington at WAC 173-303-17-202 would be

complied with for accumulating or managing such waste on-site for up to 90 days. Unanticipated circumstances could require

compliance with other hazardous/dangerous waste requirements. State dangerous waste is defined more broadly than Federal

hazardous waste.

Land Disposal of Waste Management and disposal of materials

containing polychlorinated biphenyls (PCBs)

Toxic Substances Control Act (15 U.S.C. §

2605; 40 CFR 761.61(c))

Dangerous Waste Management (RCW

70.105; WAC 173-303- 140, 141)

Based on the RI, dredged materials with PCB remediation waste as defined in 40 CFR 761.3 is not anticipated. Any such

dredged material will be subject to EPA-approved plans for all cleanup activities, including any sampling, as well as all on-site

disposal-related activities. Risk based disposal of PCB remediation wastes must not pose unreasonable risk of injury to health

or the environment. Written EPA approval is required for any PCB remediation waste off-site disposal.

Hazardous waste Resource Conservation and Recovery Act

Land Disposal Restrictions (42 U.S.C. §§

6901-6992K; 40 CFR 268)

See Dangerous or Hazardous Waste Management project-specific coments above. Any dangerous or hazardous waste land

disposal shall meet substantive land disposal requirements.

Dredge/Fill and Other

In-Water Construction

Work

Discharge of dredged/fill material into

navigable waters or wetlands

Clean Water Act Sections 401, 404 (33

U.S.C. §§ 1341, 1344; 40 CFR 121.2

(content of 401 certifications), 230 (disposal

sites/mitigation), 232

(definitions/exemptions); 33 CFR 320, 322-

3, 328-30 (Army Corps of Engineers 404

Permitting))

Hydraulic Code Rules

(RCW 77.65; WAC 220-110)

Dredged Materials Management Program

(DMMP) (RCW 79.105.500; WAC 332-30-

166 (3))

401: EPA will issue the equivalent of state certification assuring water quality standards will not be violated by remedial action

discharges along with necessary conditions including any mixing zone parameters consistent with WAC 173-201A-400, as

developed in remedial design.

404: Substantive dredge or fill criteria and requirements for discharges will be met, along with substantive mitigation

requirements for unavoidable loss of aquatic habitat; mitigation will be assessed and defined as necessary in remedial design.

Hydraulic codes provide construction criteria, requirements and limitations, including for dredging, piers, piles, docks, bulkheads

and bank protection, specified technical provisions, special concerns.

The use of an established open-water disposal site for dredged material for which there is no practical alternative upland

disposal site or beneficial use as set forth in WAC 332-30-166(3) will be approved by the designated federal and state DMMP

agencies.

Navigation and commerce Rivers and Harbor Act Section 10

(33 U.S.C. § 403)

Unauthorized obstruction or alteration of navigable waterways is prohibited. Dredging/capping residual elevations will be

designed to preserve navigation and commerce. In-water disposal is not anticipated; any in-water disposal site will not obstruct

or alter navigation upon completion.


108

Topic Standard or Requirement

Regulatory Citation

Project-Specific Comments Federal State

Endangered Species and

Critical Habitat

Taking or jeopardy to endangered or

threatened species; adverse modification of

critical habitat

Endangered Species Act (16 U.S.C. §§

1531-1544; 50 CFR 17 (listings,

prohibitions), 402 (interagency

consultations), 222-224 (endangered and

threatened marine species), 226.212 (critical

habitat for Northwest salmon and

steelhead))

It is unlawful to take (or possess, deliver, carry, transport or ship) any endangered species, or violate any regulation

(promulgated pursuant to Section 4) re endangered or threatened species. EPA in consultation with the Services shall insure

any authorized action is not likely to jeopardize endangered or threatened species or adversely modify critical habitat, absent an

exemption. EPA shall prepare a Biological Assessment for the Services which will produce a Biological Opinion including any

reasonable and prudent alternatives or measures to be taken which will guide remedy implementation, including within specified

time periods (“fish windows”) for specified activities.

Migratory Birds Taking or adversely affecting migratory

birds.

Migratory Bird Treaty Act, (16 U.S.C §§ 703-

712; 50 CFR 10 and 21)

Remedy will be carried out in a manner to avoid adversely affecting migratory bird species as defined in federal regulations,

including individual birds and their nests.

Eagles Taking or harming eagles Bald and Golden Eagle Protection Act (16

U.S.C. § 668, 50 CFR 22)

Bald Eagle Protection Rules (RCW

77.12.655; WAC 232-12-292)

Taking or harming of eagles, their eggs, nests or young is prohibited; substantive requirements for the protection of bald eagle

habitat including nesting, perching and roosting sites will be met.

Floodplain Protection Adverse impacts; potential harm Floodplain Management Procedures (40

CFR 6, Appendix A, Section 6, see also

Executive Order 11988)

The required evaluation of potential effects of authorized remedial action, to avoid adverse impacts and to minimize impacts for

which no practicable alternative exists, followed as necessary by the development of avoidance and/or minimization plans, will

be undertaken during remedial design.

Shoreline management Construction and development Shoreline Management Act RCW 90.58;

WAC 173-26; City of Seattle Master Plan

SMC 23.60;King County Master Plan

K.C.C. 21A.25)

Master plans within their jurisdiction apply within 200 feet of the shoreline to the extent they impose or establish more stringent

requirements. Compliance as may be necessary will be evaluated during remedial design.

Air Emissions Ambient air quality standards; fugitive

emission/fugitive dust

Clean Air Act (42 U.S.C. §§ 7401-7671q; 40

CFR 50)

Washington Clean Air Act (RCW 70.94;

WAC 173-400)

Any source of fugitive emissions or fugitive dust must take reasonable precautions to 1) prevent the release of air contaminants,

2) prevent fugitive dust from becoming airborne, and 3) maintain and operate the source to minimize emissions. See especially

WAC 173-400-040(4) and (9).

Native American Graves

and Sacred Sites

Protections Native American Graves Protection and

Repatriation Act (25 U.S.C. §§ 3001 et

seq.); American Indian Religious Freedom

Act (42 U.S.C. §§ 1196 et seq.)

Requirements for the protection of Native American remains, funerary objects and associated cultural artifacts when burial sites

are encountered; and protection of tribal exercise of traditional tribal religions, including traditional cultural properties, sites and

archeological resources. See also Executive Order 13007 which requires federal agencies to avoid physical damage to tribal

sacred sites, and interfering with access of tribes thereto. Compliance will be maintained throughout remedy implementation as

may be necessary

Noise Permissible noise levels Noise Control Act (RCW 70.107; WAC

173-60-040-050)

Maximum levels at specified times for specified durations are in 173-60-040, subject to exemptions in 173-60-050, including

050(3)(a) (sounds originating from temporary construction sites as a result of construction activity) and (3)(f) (sounds created by

emergency equipment and work necessary in the interests of law enforcement or for health, safety or welfare of the community).

Historic Preservation National Historic Preservation Act Section

106 (16 U.S.C. § 470; 36 CFR 800)

The effect if any of remedial activity on any district, site, building, structure or object included or eligible for inclusion in the

National Register of Historic Places will be evaluated in consultation with the State Historic Preservation Office during remedial

design.


109

10.2 Balancing Criteria

The balancing criteria evaluate the major trade-offs among alternatives.

10.2.1 Long-Term Effectiveness and Permanence

Although all alternatives are predicted by RI/FS models to result in the same long-term risks after

cleanup, as discussed in Section 10.1.1, alternatives that rely less on natural recovery have less

uncertainty in long-term model projections, though this uncertainty can be reduced to some extent

through monitoring and adaptive management, as is required for MNR To Benthic SCO. The higher-

numbered alternatives have increasingly larger cleanup footprints, and rely progressively less on natural

recovery to achieve cleanup objectives.

Alternatives that remove more surface and subsurface contamination through dredging provide the most

permanence, followed by those that effectively isolate it through engineered caps. Dredged contaminated

sediment is permanently removed from the LDW, and capped sediment is securely segregated from

contact with receptors. Caps typically maintain their effectiveness as long as they are monitored and

maintained. Contamination remaining in subsurface sediments and not isolated by a cap would contribute

to future risks if it were brought to the surface of the waterway through natural or man-made events such

as earthquakes, vessel scour, or construction activities. ENR and MNR are not designed to isolate

contamination, so alternatives that use these in areas with lower contaminant concentrations provide

better long-term effectiveness than those that use them in areas with higher concentrations. The potential

to increase surface sediment COC concentrations through disturbance of subsurface sediments is not

accounted for in the BCM; thus, the BCM may underestimate the long-term COC concentrations.

All alternatives would remediate sediments with contaminant concentrations exceeding RALs in the top

60 cm in Recovery Category 1 areas, to address the potential for exposure of subsurface contamination in

the areas where disturbance is most likely. All alternatives also remediate sediments with contaminant

concentrations exceeding the direct contact (RAO 2) RALs for arsenic, cPAHs, and dioxins/furans in the

top 45 cm in intertidal sediments to protect people clamming or digging on the beach. However, it is not

possible to anticipate every location where disturbance might occur. The Selected Remedy adds

remediation of sediments with contaminant concentrations exceeding PCB RALs of 65 mg/kg OC in the

top 45 cm in intertidal sediments to reduce the potential for exposure of subsurface contamination through

digging on the beach, and 195 mg/kg OC in the top 60 cm of subtidal sediments to reduce the influence of

activities such as vessel scour.21

It also addresses the potential for release of subsurface sediment

contamination through navigation dredging. Alternatives 5R, 6C, and 6R would remediate more

subsurface sediment contamination in all potentially erosive areas than the Selected Remedy.

21

LDWG (2012b) evaluated several options for subsurface sediment RALs. EPA selected the 5C Plus RALs listed

here because other options would remove less subsurface contamination with an associated increased risk of

exposure, or would remove more subsurface contamination but at a higher cost that was disproportional to the

increase in long-term effectiveness and permanence. The PCB RAL of 195 mg/kg OC is applied in the top 60 cm

(potential vessel scour depths) in Recovery Categories 2 and 3; see Table 22, footnote a and Section 13.1. In the

intertidal zone for Alternative 5C Plus, all intertidal RALs must be met to 45 cm depth in Recovery Category 1

areas instead of the top 60 cm for FS alternatives. This is because 45 cm was deemed by EPA to be sufficiently

protective in intertidal areas.


110

Monitoring of sediment, fish and shellfish tissue, and surface water would be required under all remedial

alternatives. Areas that are dredged would require the least long-term sediment monitoring. Capping and

ENR require more long-term sediment monitoring to ensure surface sediment concentrations remain low.

Areas with MNR would require the most monitoring to determine if surface sediment COC

concentrations are reducing over time as projected by the natural recovery model. Alternatives with a

larger area of ENR and MNR require more long-term monitoring and maintenance to ensure their

effectiveness than dredged areas would.

All alternatives rely on institutional controls (ICs) to reduce exposure to contamination remaining after

remediation. Alternatives that rely more on removal of contamination from the waterway through

dredging rely less on institutional controls. Resident seafood consumption advisories were included for all

alternatives; these are informational devices that have historically had limited effectiveness according to

published studies and in EPA’s experience. As noted in EPA’s EJ Analysis, the community and affected

Tribes have identified several concerns about the use of ICs, including the burden placed on Tribes

exercising their treaty rights and on other fish consumers in the LDW as the traditional ICs assume there

are accessible substitute food sources and that changing behavior is appropriate and acceptable. To

address this concern to the extent practicable, the Selected Remedy adopts the EJ Analysis

recommendation that the affected community and Tribes with treaty rights be directly involved in

advising EPA on institutional controls development, and that enhanced outreach and education programs

be developed. These outreach efforts include periodic seafood consumption surveys to identify what

species are being eaten by whom, as well as understanding the perceptions of risks and benefits of

consuming fish from the LDW. The first seafood consumption survey (Fishers Study; LDWG 2014b) is

currently underway, with many opportunities for meaningful involvement by the affected fish consuming

community, as recommended by the National Environmental Justice Advisory Council (NEJAC). EPA

anticipates that the information collected through the Fishers Study may serve as a basis for developing

more effective and appropriate ICs, and a more targeted education and outreach program.

Other ICs such as environmental covenants or restricted navigation areas were included to protect caps.

Institutional controls in either of these forms regarding ship/vessel use and/or restrictions on anchoring or

spudding (sinking vessel-mounted poles into sediment for stabilizing vessels) would have been used

under all alternatives to reduce the possibility of releases of COCs in underlying sediments. However,

these restrictions may have limited reliability in much of this heavily used waterway.

Alternative 2R-CAD would have the least long-term effectiveness and permanence because it leaves the

largest amount of contamination in place. It also would require long-term maintenance of a CAD site in

the LDW. Similar to caps, CADs typically maintain their effectiveness as long as they are monitored and

maintained. Alternatives that would leave progressively less subsurface contamination in the waterway

are progressively more effective for this criterion. The Selected Remedy (5C Plus), 5R, 6C,and 6R would

be the most effective, while other alternatives would be comparatively less effective based on the amount

of subsurface contamination left behind.

10.2.2 Reduction of Toxicity, Mobility, or Volume through Treatment

Alternative 5R-Treatment would utilize ex situ soil washing to reduce volumes that would be disposed in

a landfill. For all alternatives that include ENR, the FS assumed that 50% of the ENR area would include

in situ treatment (e.g., use of activated carbon or other amendments) to reduce toxicity and bioavailability.

Thus, the reliance on in situ treatment would be proportionate to the amount of ENR. Alternative 6C


111

(with 101 total ENR acres) would have the greatest reliance on ENR (with potential in situ treatment);

while Alternatives 4C (with 16 total ENR acres), and 3C (with 10 total ENR acres) would utilize this

technology significantly less. The Selected Remedy is in the mid-range, with 48 total ENR acres. Both in

situ and ex situ treatments would require verification and bench or pilot scale testing during the remedial

design phase.

Principal threat waste is defined in EPA guidance as source material that is highly toxic or highly mobile,

such as pools of non-aqueous phase liquids, and that generally cannot be reliably contained or would

present a significant risk to human health or the environment should exposure occur; see Section 11. No

direct evidence of any significant amounts of non-aqueous phase liquids has been found in LDW

sediment; however, treatment was included for some alternatives.

Based on these considerations, Alternative 5R-T would have provided the most treatment of contaminated

sediment. The “C” alternatives, including 5C Plus (the Selected Remedy), would also provide treatment,

the degree of which is based on the amount of area identified for ENR/in situ treatment.

10.2.3 Short-Term Effectiveness

Short-term impacts associated with the cleanup alternatives may include traffic, noise, air emissions,

habitat disturbance, and elevated fish tissue concentrations during implementation of the cleanup. Local

transportation impacts (traffic, noise, air pollution) from implementation of these alternatives are

proportional to the amount of dredging or the amount of capping, fill, and ENR materials that would have

to be transported. Among the technologies evaluated, dredging has the highest potential for short-term

impacts because it takes longer to implement than other technologies, requires transportation of sediments

to a landfill, and creates more disturbance of contaminated subsurface sediments. Short-term impacts

identified in the RI/FS will be evaluated further during remedial design and efforts will be made to

mitigate them and otherwise enhance the environmental benefits of the remedy consistent with CERCLA,

the NCP, and EPA Region 10's Green Remediation policy. For example, impacts due to construction

would be reduced to the extent possible using best management practices and performing in-waterway

work only when threatened juvenile salmon are not migrating through the waterway. Dredging often

leaves residual contamination behind; this would be managed by placing a thin layer of clean sand in

areas where RALs are not met after dredging.

The Selected Remedy adopts recommendations from EPA's EJ Analysis (Section 10.3.3) to reduce risks

as quickly as possible to minimize impacts to the community and Tribes during construction. As

recommended by the EJ Analysis, EPA would work with the community and consult with Tribal

representatives to reduce impacts on community and Tribal resources during remedy implementation.

Figure 14 summarizes the construction time and predicted time for each alternative to achieve modeled

risk reduction benchmarks associated with each RAO. These estimates were derived from the estimated

time to complete construction and the estimated natural recovery periods predicted by the BCM. As

discussed in Section 10.1.1, it was not possible to predict the time to achieve all cleanup levels. Generally,

the potential for short-term impacts increases as the length of the construction period (based on the area

and volume to be actively remediated) increases. Lower-numbered alternatives, with higher RALs, could

be implemented faster than higher-numbered alternatives with lower RALs. However, lower-numbered

alternatives also relied more on natural recovery and therefore had more uncertainty in their long-term

effectiveness.


112

Alternatives 5C and 5C Plus (the Selected Remedy) provided the best short-term effectiveness by

providing the best balance of a relatively short construction time (7 years) along with a relatively short

projected time to reach all FS cleanup objectives (17 years). While Alternatives 2R, 3C, 3R, and 4C had

shorter construction times, (3 – 6 years), they were projected to take longer to reach all FS cleanup

objectives (18 – 24 years) because of the extended time needed for natural recovery. Alternative 6C was

projected to take 16 years to reach all FS cleanup objectives, but had a much longer construction time of

16 years. Alternative 6R had a long construction time and time to reach FS cleanup objectives of 42 years.

10.2.4 Implementability

Technologies used in these cleanup alternatives have been implemented successfully at other projects in

the Puget Sound region. Alternatives with longer construction times and lower RALs have the potential

for more delays or difficulties. The use of in situ treatment technologies in association with ENR is a

relatively new technology in the Puget Sound region and will require pilot testing before full

implementation. The soil washing component of Alternative 5R-Treatment had potential technical and

administrative challenges associated with locating and permitting an upland soil washing facility, and

potentially with reuse or disposal of treated material. Treatability studies would have been required to

verify the suitability of soil washing as a viable treatment technology.

Alternatives with higher RALs and larger MNR footprints have a higher potential for requiring additional

actions if MNR To Benthic SCO does not reduce contaminant concentrations as expected. This may have

caused an additional administrative burden to determine specific additional actions, and to provide

oversight during implementation of such actions. Alternative 2R-CAD had a potentially significant

administrative challenge related to locating, using, and maintaining one or more CAD facilities.

Institutional controls were a requirement of all remedial alternatives to manage human health risks from

consumption of resident seafood. The primary control mechanisms are seafood consumption advisories,

public education and outreach, governmental controls, and environmental covenants pursuant to the

Washington Uniform Environmental Covenants Act, which can be difficult to monitor. Seafood

consumption advisories are not enforceable, and have limited effectiveness. For these reasons,

alternatives that rely less on institutional controls are more readily implementable.

Alternatives 5R, 5R-T, 6R, and 2R-CAD had the greatest potential implementability challenges due to the

long construction timeframes for 5R and 6R and the difficulties associated with building and operating a

soil washing technology (5R-T) or in finding suitable location(s) and meeting substantive permit

requirements for a CAD(s) with sufficient capacity (2R-CAD). Alternatives 4C, 4R, 5C, and 5C Plus (the

Selected Remedy) are all similarly highly implementable because they rely on technologies that have

been proven effective at other cleanups and are administratively feasible, and their large actively

remediated areas equate to a low probability for triggering additional actions in the future.

10.2.5 Cost

Capital, operations, maintenance and monitoring (OM&M), and 30-year net-present value (NPV) costs

for each alternative, calculated with a 2.3% discount rate, are provided in Table 25. Table 25 also shows

NPV costs using a range of discount rates, from 0% to 7%. The estimated cost of $342 million (NPV at

2.3% discount rate) for the Selected Remedy falls within the low end of the cost range for the FS

alternatives ($200 million – $810 million).


113

10.3 Modifying Criteria Consideration of Modifying Criteria was based on the public comments received on the Proposed Plan,

discussions with the State, and consultation with affected Tribes. The findings and recommendations in

and comments on EPA’s EJ Analysis were also considered as part of the two modifying criteria.

10.3.1 Community Acceptance

EPA received 2,327 comments on the Proposed Plan from individuals, businesses, interest groups, Tribes,

and government agencies. Comments on the extent of cleanup were sharply divided. In general,

community groups and individuals wanted stringent cleanup levels, a more extensive cleanup than the

Preferred Remedy in the Proposed Plan, and a cleanup that would provide for safe consumption of

resident fish and shellfish without fish advisories. The parties who conducted the RI/FS and some

businesses in general wanted less stringent cleanup levels for sediments, no cleanup levels for tissue or

surface water, and a less extensive cleanup. Businesses also requested that there be flexibility in decision-

making. A summary of comments received and EPA's responses are provided in Part 3 of this ROD, the

Responsiveness Summary. Community concerns were also considered during the development of and

after review of public comments on EPA's EJ Analysis, discussed below.

10.3.2 State/Tribal Acceptance The Washington State Department of Ecology concurs with the Selected Remedy. A copy of their

concurrence letter is provided as Attachment 1.



manages aquatic resources north of the Spokane Street Bridge, just north of the LDW study area. In their

comments on the Proposed Plan, both Tribes emphasized the importance of preserving Tribal treaty-

reserved resources, and that the cleanup must be adequate to protect the health of tribal fishers exercising

their treaty rights in this area and for the protection of the aquatic ecosystem, which contributes to the

health of the fishery itself. They also stated that they have been working cooperatively with EPA since the

onset of the RI/FS and plan to continue doing so throughout the life of the project.

10.3.3 Environmental Justice Analysis

EPA’s EJ Analysis examined the impacts of the Proposed Plan Preferred Alternative and other FS

alternatives on those who subsist on, work in, and play in the LDW. The EJ Analysis considered the

potential for disproportionate adverse impacts from the FS alternatives and the Selected Remedy on the

community and affected Tribes, particularly those who consume resident fish and shellfish or have

contact with LDW sediments. In the EJ Analysis, the cleanup alternatives were compared qualitatively for

their long-term and short-term residual cancer and noncancer risks; the time to achieve human health

targets; the certainty of the methods used to conduct the cleanup; and the dependence upon institutional

controls which have implications for environmental justice concerns on behalf of the affected community.

EPA’s EJ Analysis recommended additional measures to mitigate disproportionate adverse impacts. The

EJ Analysis lists the following recommendations:

1. Emphasize reduction of greatest human health risks as soon as possible while ensuring that cleanup

methods used will be effective and last over the long-term;

2. Form and fund an advisory group with support for local community outreach experts to meaningfully

involve the community in developing the most appropriate mitigations for exposure from eating

resident seafood at the Site;


114

3. Continue support for Tribal consultation, participation, and early involvement;

4. Support a local fisher consumption survey specific to the LDW (to find out where, when, and what

they are fishing for to provide critical information in the development of institutional controls, offsets,

and enhanced education)22

;

5. Establish a mechanism to provide offsets in the event of higher short-term concentrations in fish

tissue in the LDW: fish trading may be the most straightforward, but there might be cost savings

through a sustainable aquaculture or alternative transportation method;

6. Use green remediation techniques, such as technologies that reduce air impacts, with any cleanup

alternative chosen.

The findings and recommendations of the EJ Analysis were incorporated into the evaluation of

alternatives and the Selected Remedy, except that EPA did not select any of the offsets described in item

5 as part of the Selected Remedy, as discussed in Section 13.

10.4 Summary of CERCLA Nine-Criteria Evaluation

Alternative 1 was not protective; it was therefore eliminated from further consideration. Alternatives 2 to

6 rely on MNR to varying degrees. They are potentially protective and are projected by RI/FS models to

provide substantial risk reduction. All alternatives also rely on seafood consumption advisories as

institutional controls to limit seafood consumption to protect human health. These alternatives may also

meet ARARs, although whether the Selected Remedy or any of the alternatives, in conjunction with

source control, can achieve all ARARs-based human health cleanup levels and water quality ARARs is

uncertain. As discussed Section 10.1.2, any of the alternatives may require ARAR waiver(s) in a future

ROD Amendment or ESD if ARARs are not achieved.

Alternative 2R-CAD provided the least long-term effectiveness and permanence because it would require

long-term maintenance of a CAD site within the waterway and would leave the most subsurface

contamination in place. The removal-emphasis alternatives, 2R through 6R, would leave progressively

less subsurface contamination in place that could be exposed by vessel scour or earthquakes, and would

require fewer use restrictions and less maintenance. They also would have comparatively longer

construction times and are more expensive than combined (“C”) alternatives with similar RALs. The

combined alternatives, 3C through 6C, and especially the lower-numbered combined alternatives, would

have more area managed by ENR and MNR (and thus more subsurface contamination left in place), and

would have greater monitoring and maintenance requirements. The Selected Remedy adds to Alternative

5C an increased emphasis on removal of high levels of PCBs in shallow subsurface sediments, providing

greater permanence because it addresses potential vessel scour, emergency construction, navigation

dredging, and other activities that could cause releases of subsurface contamination to the surface and

water column. Under the Selected Remedy, MNR is selected only in areas with moderate to low

concentrations of non-human health COCs to allow for a moderate construction time and to shorten the

overall time to achieve FS cleanup objectives. While Alternatives 5R, 6C and 6R would remove more

subsurface contamination, they would disrupt the waterway through releases of contamination dredging

over a much longer construction period, and considerably higher cost than the Selected Remedy (Table

24).

22.

EPA has already started to implement this recommendation as part of the RI/FS.


115

11 Principal Threat Waste The NCP establishes the expectation that treatment will be used to address the principal threats posed by a

site whenever practicable (40 CFR 300.430[a] [1] [iii] [A]). In general, principal threat wastes are those

source materials considered to be highly toxic or highly mobile that generally cannot be contained in a

reliable manner, or will present a significant risk to human health or the environment should exposure

occur.

EPA has determined that the contaminated sediments in the LDW outside of the EAAs are not highly

mobile or highly toxic. No direct evidence of any significant amounts of non-aqueous phase liquids has

been found in LDW sediments. The maximum concentrations detected for the four human health risk

drivers in surface and subsurface sediment outside of the EAAs are:

11,000 µg TEQ/kg dw for cPAHs

2,100 ng TEQ/kg dw for dioxins/furans

890,000 µg/kg dw for total PCBs

2,000 mg/kg dw for arsenic

Direct contact risks are low relative to seafood consumption risks (maximum direct contact RME excess

cancer risk is 2 × 10-4

, as compared to an excess cancer risk of 3 × 10-3

for seafood consumption). For

PCBs and dioxins/furans, the primary threat comes from bioaccumulation through exposure of aquatic

receptors (e.g., fish and shellfish). Once contaminated sediment is capped or dredged, exposure through

seafood consumption will cease.

Most alternatives, including the Selected Remedy, would utilize ENR/in situ treatment if pilot testing

shows that the technology will be effective.


116

12 Documentation of Significant Changes to the Selected Remedy

In response to comments received on the Proposed Plan, EPA has altered some aspects of the Preferred

Alternative (Alternative 5C Plus) in the Proposed Plan in formulating the Selected Remedy. This section

briefly describes the changes, which are discussed in more detail in Sections 8 and 13 and in Part 3, the

Responsiveness Summary.

Cleanup Levels. EPA received many comments on the PRGs in the Proposed Plan, as described

in Part 3. In response to those comments and as discussed in Section 8:

o The sediment PRGs in the Proposed Plan are cleanup levels in the ROD though they were

derived differently due to revisions to the SMS as discussed in Section 8.2 and 14.2.

o The ambient water quality criterion (AWQC) for PCBs was a proposed ARAR and PRG

in the Proposed Plan. For all other COCs, the AWQCs were identified as ARARs but not

as cleanup levels. As indicated in the Proposed Plan, PCBs are the most widely

distributed of the LDW COCs and may have merited special consideration. After

considering comments on the Proposed Plan, as described further in the Responsiveness

Summary (Part 3 of this ROD), EPA has decided to address PCBs in the ROD as it did all

other COCs in LDW surface water. As a result, the remedial action selected in this ROD

identifies the AWQCs (or State WQS or EPA's 1992 National Toxics Rule (NTR)

standards, whichever is lower; see Section 8.2.2.2) for all COCs, including PCBs, as

ARARs. EPA recognizes that achieving water quality ARARs in the In-waterway

Portion of the LDW in part depends on source control actions to be taken under other

authorities in addition to the CERCLA remedial action selected in this ROD. Consistent

with 40 CFR 300.430(f), this ROD identifies AWQC as ARARs when they are lower

than State WQS or NTR, and EPA intends this final remedial action will attain the

AWQC, consistent with 40 CFR 300.430(f). See further discussion of surface water

quality ARARs in Section 14.2.

o The fish and shellfish tissue PRGs in the Proposed Plan are called target tissue

concentrations in the ROD, and will only be used to assess remedy implementation [40

CFR 300.430(f)(5)(iii)]. This ROD establishes no tissue cleanup levels. EPA decided to

use these levels as a means of assessing the success of the Selected Remedy in

conjunction with State-led source control. EPA proposed tissue PRGs in the Proposed

Plan because people consume fish and shellfish tissue, therefore tissue concentrations are

the best and most direct measure of risk to LDW resident fish and shellfish consumers

from those COCs. However, fish and shellfish derive their COC concentrations from both

sediments and surface water in proportions that at this time can only be approximated and

estimates of the degree to which this CERCLA action will reduce fish and shellfish tissue

concentrations are highly uncertain. Therefore, EPA decided to identify target fish and

shellfish tissue concentrations rather than cleanup levels.

Remedial Action Levels. For the Selected Remedy, benthic protection RALs are 2 times the

benthic SCO criteria in Recovery Category 2 and 3 areas; they were 2 times the benthic SCO but


117

not to exceed the CSL criteria in the Proposed Plan. This change was introduced because many

COCs have CSLs that are close to or equal to the SCO, and EPA has determined that setting the

RAL at twice the SCO is sufficiently protective of benthic invertebrates during the natural

recovery period. This RAL change does not affect the cleanup levels, which remain as they were

in the Proposed Plan; the SCO criteria must be met no later than 10 years following construction

completion. See Section 13.2.

Buffer Depth Above Caps Placed in the Federal Navigation Channel. The required depth,

below the authorized channel depth, for the top of any cap in the federal navigation channel is

increased from 3 ft (which was used in the FS) to 4 ft to address concerns raised in comments

from the US Army Corps of Engineers (USACE 2013). USACE typically “advance dredges” 2 ft

below the authorized depth so that sedimentation in the channel does not interfere with navigation

for a longer period and dredging is required less frequently. An additional 2 ft is needed to ensure

that navigation dredging does not damage the cap. See Section 13.2.1.

Updated Dredging Volume and Cost Estimates. EPA received several comments on the

Proposed Plan suggesting that EPA update volume and cost estimates using data collected after

completion of the FS. The Selected Remedy addresses these comments by updating estimated

dredged volumes and costs using data from a 2012 USACE characterization of subsurface

sediments in shoaled areas in the navigation channel (USACE 2013b). Although other studies

have been conducted after completion of the FS, EPA determined that this study would likely

have the most significant impact on volume and cost estimates. See Sections 13.2.1 and 13.3.

Dredge Contaminated Shoaled Areas in the Navigation Channel. EPA received several

comments from USACE and waterway users about the importance of maintaining the LDW

navigation channel at its authorized depth. USACE commented that the Preferred Alternative in

the Proposed Plan, by proposing to dredge shoaled areas in the navigation channel only if

contaminant concentrations in the top 2 ft exceed RALs, did not sufficiently address

contamination in the navigation channel and would impede their ability to maintain the

navigation channel. EPA agrees that contaminated sediments in the navigation channel should be

removed to provide a remedy that is compatible with current and reasonably anticipated future

use of the waterway. The Selected Remedy addresses this concern by requiring that all sediments

in the navigation channel be dredged if COC concentrations exceed RALs at any depth above the

maintenance depth (2 ft below the authorized depth). See Section 13.2.1.

Reevaluate Recovery Categories. EPA received many comments suggesting that Recovery

Categories were incorrectly assigned to specific locations and asking EPA to reevaluate Recovery

Categories based on new information collected after the RI/FS was completed. EPA will

reevaluate assignment of specific areas to Recovery Categories using information collected after

the RI/FS and during remedial design as discussed in Section 13.2.3.

Flexibility in Assignment of Cleanup Technologies and Institutional Controls. Many

waterway users and community members emphasized in their comments the importance of

maintaining the existing uses of specific areas of the waterway. As part of remedial design,


118

surveys will be conducted to better understand the operational areas and needs of waterway users,

and EPA will work with users to minimize conflicts between use restrictions needed to maintain

the integrity of caps and other remedy features in light of existing and reasonably anticipated uses

of the LDW, as discussed in Section 13.2.4. This may require implementation of dredging in

some areas that may otherwise be eligible for capping because capping would require use

restrictions that are incompatible with waterway uses.

The changes from the estimate in the Proposed Plan due to modifying the area and volume requiring

dredging and cost of the Selected Remedy are as follows:

Incorporating new data and increasing depth requirements for caps in the navigation channel by

1 ft adds 9 acres, 70,000 cy, of sediments to be dredged, and $18 million.

Dredging all contaminated sediments in shoaled areas in the navigation channel adds 12 acres,

100,000 cy, of sediment to be dredged, and $27 million. This element can be implemented as part

of remedial action or phased in as USACE identifies the need for dredging to address

impediments to navigation.

Total additions: 21 acres, 170,000 cy, and $45 million.

All these estimates are based on limited data and will be refined during remedial design. All changes are

within the expected accuracy of FS cost estimates of +50% to -30%, and are less than the 200,000 cy

sediment to be dredged that was used as a contingency volume in the FS cost estimates. They are within

the range of adjustments that would normally be made during remedial design, and do not significantly

change the Selected Remedy.


119

13 Selected Remedy Based on consideration of the requirements of CERCLA, the detailed analysis of remedial alternatives,

and consideration of public comments, EPA has selected Alternative 5C Plus as the Selected Remedy,

with modifications summarized in Section 12, for the In-waterway Portion of the Site. This section

provides EPA’s rationale for the Selected Remedy, and a description of its anticipated scope, how the

remedy will be implemented, and its expected outcomes.

13.1 Summary of the Rationale for the Selected Remedy

The Selected Remedy is protective of human health and the environment, complies with ARARs, and

provides the best balance of tradeoffs among the balancing criteria. It reduces risks within a reasonable

time frame, is practicable and cost-effective, provides for long-term reliability of the remedy, and

minimizes reliance on institutional controls. It will achieve substantial risk reduction by dredging and

capping the most contaminated sediments, reduce remaining risks to the extent practicable through ENR

and MNR, and manage remaining risks to human health through institutional controls.

EPA considered several options for surface sediment and subsurface sediment RALs that determine

where active (dredging, capping, ENR) and passive (MNR) response actions will be applied. EPA

selected the RALs listed in this ROD because alternatives with higher RALs would remove less

subsurface contamination, resulting in less certainty in achieving cleanup goals. Alternatives with lower

RALs and more emphasis on dredging would remove more subsurface contamination at higher cost and

potentially greater short-term risks, with uncertain associated increases in long-term protectiveness. More

than other alternatives, the Selected Remedy emphasizes a combined-technology approach, including

removal of shallow subsurface sediments with higher concentrations of PCBs, while allowing MNR in

areas with lower concentrations of other COCs. The Selected Remedy provides better long-term

effectiveness than other alternatives by adding remediation of sediments in subtidal areas with high

concentrations of PCBs in the top 2 ft below the surface of the sediment in Recovery Category 2 and 3

areas, whereas other alternatives propose remediation of subtidal contamination in the top 2 ft below the

sediment surface only in Recovery Category 1 areas. This addition provides better protection for releases

of contamination that may occur due to infrequent events, such as vessels traveling outside of frequent

lanes of operation, vessels operating with excessive propeller power in berthing areas or elsewhere, barge

groundings, emergency maneuverings, changes in the patterns of site use, and maintenance of overwater

structures. It also addresses contamination in sediments in the navigation channel that may otherwise be

released during maintenance dredging.

For all of these factors, the Selected Remedy provides greater permanence in comparison to other

alternatives of similar cost and construction duration. Less costly alternatives rely on technologies such as

ENR and MNR to address areas with higher COC concentrations, resulting in greater uncertainty as to

their long-term effectiveness. In more costly alternatives, the additional costs are not proportional to the

overall increase in long-term effectiveness.

The Selected Remedy provides the best balance of minimizing short-term risks due to a comparatively

short 7-year construction period, while maximizing long-term effectiveness by dredging or capping the

most contaminated sediments. The Selected Remedy will utilize treatment to reduce the toxicity and

bioavailability of contaminants in the form of ENR with in situ amendments if pilot testing is successful.


120

13.2 Description of the Selected Remedy

The Selected Remedy addresses all areas where contaminant concentrations exceed the cleanup levels

through a combination of active cleanup technologies, monitored natural recovery, and institutional

controls. See Section 8 for a discussion of cleanup levels. The approximate areas that would be

remediated through dredging, partial-dredging and capping, capping, or ENR and ENR/in situ treatment,

and areas where COC concentrations would be reduced through MNR both above and below the benthic

SCO, are shown in Figure 18 on page 137.

In summary, the Selected Remedy consists of the following elements:

Apply active cleanup technologies in a total of 177 acres, as described in Figure 19 and Figure 20:

Dredge or partially-dredge and cap approximately 105 acres of highly contaminated sediments

(approximately 960,000 cubic yards).

Place engineered sediment caps on approximately 24 acres of highly contaminated sediments

where there is sufficient water depth for a cap.

Place a thin layer (6 to 9 inches) of clean material (referred to as enhanced natural recovery

[ENR]) on approximately 48 acres of sediments in areas that meet the criteria for ENR.

Apply location-specific cleanup technologies to areas with structural or access restrictions (e.g.,

under-pier areas and in the vicinity of dolphins/pilings, bulkheads, and riprapped or engineered

shorelines).

Implement monitored natural recovery (MNR) in approximately 235 acres of sediments where

surface sediment contaminant concentrations are predicted to be reduced over time through deposition of

cleaner sediments from upstream. MNR will apply to those areas that are not subject to active

remediation, using either MNR To Benthic SCO or MNR Below Benthic SCO, as described in Section

13.2.2 and in Figure 21.

Sample the entire LDW (441 acres) as part of baseline, construction, post-construction, and long-

term monitoring. Conduct sampling and analysis to establish post-EAA cleanup baseline conditions

during remedial design, and conduct construction, post-construction, and long-term monitoring, as

described in Section 13.2.3.

Provide effective and appropriate institutional controls (ICs) for the entire waterway to reduce

human exposure to contaminants, ensure remedy protectiveness, and protect the integrity of the remedy,

while minimizing reliance on ICs, particularly seafood consumption-related ICs, to the extent practicable,

as described in Section 13.2.4.

The estimates of areas, volumes, time to reach cleanup objectives, and cost for the Selected Remedy in

this ROD are based on RI/FS data and other information included in the Administrative Record. Remedial

design sampling will be conducted after cleanups are completed in the Early Action Areas. Results from

remedial design sampling will be used to refine delineation of areas to be remediated by varying

remediation technologies and the remediation technologies to be applied, and inform source control

activities. This section describes how data collected in the future will be used to revise the delineation of

areas requiring cleanup and the technologies applied to each area.


121

13.2.1 Application of Cleanup Technologies

The RALs listed in Figure 22 and Figure 23 (above) and Table 27 and Table 28 (page 125)will be applied

in intertidal and subtidal areas in Recovery Category Areas 1, 2, and 3 to identify areas for active

remediation, as described and in Figures 19 and 20. Recovery Category areas are shown in Figure 12.

Figure 17 shows Recovery Category 1, and potential scour areas in Recovery Categories 2 and 3. All of

this information will be used to determine the appropriate compliance depth for application of RALs and

technology to be applied at a particular location, as described in this section.

Table 27. Selected Remedy RAO 3 RALs

SMS Contaminant of Concern for

RAO 3

RAL for Recovery Category 1

Areasa (Benthic SCO)

RAL for Recovery Category 2 & 3

Areas (2 x Benthic SCO)b

Metals (mg/kg dw)

Arsenic 57 n/a

Cadmium 5.1 10.2

Chromium 260 520

Copper 390 780

Lead 450 900

Mercury 0.41 0.82

Silver 6.1 12.2

Zinc 410 820

PAHs (mg/kg OC)

2-Methylnaphthalene 38 76

Acenaphthene 16 32

Anthracene 220 440

Benzo(a)anthracene 110 220

Benzo(a)pyrene 99 198

Benzo(g,h,i)perylene 31 62

Total benzofluoranthenes 230 4650

Chrysene 110 220

Dibenzo(a,h)anthracene 12 24

Dibenzofuran 15 30

Fluoranthene 160 320

Fluorene 23 46

Indeno(1,2,3-cd)pyrene 34 68

Naphthalene 99 198

Phenanthrene 100 200

Pyrene 1,000 2,000

Total HPAHs 960 1,920

Total LPAHs 370 740


122

SMS Contaminant of Concern for

RAO 3

RAL for Recovery Category 1

Areasa (Benthic SCO)

RAL for Recovery Category 2 & 3

Areas (2 x Benthic SCO)b

Phthalates (mg/kg OC)

Bis(2-ethylhexyl)phthalate 47 94

Butyl benzyl phthalate 4.9 9.8

Dimethyl phthalate 53 106

Chlorobenzenes (mg/kg OC)

1,2,4-Trichlorobenzene 0.81 1.62

1,2-Dichlorobenzene 2.3 4.6

1,4-Dichlorobenzene 3.1 6.2

Hexachlorobenzene 0.38 0.76

Other SVOCs and COCs, (µg/kg dw except as shown)

2,4-Dimethylphenol 29 58

4-Methylphenol 670 1,340

Benzoic acid 650 1,300

Benzyl alcohol 57 114

n-Nitrosodiphenylamine, mg/kg OC 11 22

Pentachlorophenol 360 720

Phenol 420 840

PCBs (mg/kg OC)

Total PCBs 12 n/a

Notes: General:

PCBs and arsenic are also human health COCs (see Table 28 for RALs for human health COCs), and RALs for the the human health category take precedence over RAO 3 RALs. The surface sediment (10 cm) Recovery Category 1 RALs for PCBs and arsenic are the same for human health and benthic invertebrates, but the 2 X SCO Recovery Category 2 and 3 criteria are not applicable to PCBs and arsenic. Figure 22 and Figure 23 list all RALs for human health COCs.

Table 23 describes Recovery Categories and Figure 12 shows Recovery Category areas. a. The RAL applies to the 10 cm and 45 cm depth intervals for intertidal areas and to the 10 cm and 60 cm depth intervals for subtidal

areas. See Figure 22 and Figure 23 . b. For Recovery Category 2 and 3 areas, the RAL applies to the 10 cm depth interval. See Figure 22 and Figure 23.


123

13.2.1.1 Dredging and Capping

Dredging or partially dredging and capping will be used in areas that have a potential for erosion and

where sediments are more highly contaminated (COC concentrations are higher than ENR upper limits;

see Section 13.2.1.2 and Table 28), and where it is necessary to maintain water depth for human use and

compatibility with current and reasonably anticipated future human use, or to maintain habitat, as

described below and presented as flow diagrams in Figure 19 and Figure 20. EPA will gather detailed

information during remedial design about COC concentrations, potential for scour or disturbance, and

waterway use in specific areas to determine locations for dredging, capping, and ENR. Dredging is

required under the conditions described below:

Shoaled areas in the navigation channel (where the bottom elevation is currently shallower than

the authorized navigation depth) will be dredged if COC concentrations exceed human health

RALs (for PCBs, cPAHs, arsenic or dioxins/furans) or the benthic SCO at any depth above the

maintenance depth (defined as 2 ft below the authorized depth) (Table 28). 23

The post-dredging sediment surface must not exceed human health RALs (for PCBs, cPAHs,

arsenic or dioxins/furans) or the benthic SCO. If these levels cannot be achieved through

dredging, an ENR layer will be applied to the post-dredge surface.

If the ENR upper limits are exceeded after dredging, the area must be capped. If 1 ft or less of

contamination would remain at concentrations greater than the human health RALs or the benthic

23

Shoaled areas in the navigation channel must be dredged during the implementation of the remdial action where

contaminant concentrations in the top 2 ft exceed RALs. Where contaminant concentrations exceed RALs only

at depths below the top 2 ft, cleanup may be deferred if USACE determines it is not currently an impediment to

navigation, but must be dredged in the future if USACE determines that the area has become an impediment to

navigation.

Relationship Between RALs, ENR Upper Limits, and Cleanup Levels

Remedial Action Levels (RALs) — RALs shown in Tables 27 and 28 will be used during remedial action to

delineate areas that require active remediation (dredging, capping, or ENR). Exceedances of RALs are evaluated at

each sampling station; they are not averaged over an area. RALs apply to specific locations and depths, as

described in the tables.

Enhanced Natural Recovery (ENR) Upper Limits — ENR upper limits (Table 27) are higher concentrations than

RALs. They will be used during remedial action to delineate the areas that require capping or dredging, but are not

suitable for ENR.

Cleanup Levels — Cleanup levels shown in Table 19 for RAOs 1, 2, and 4, and Table 20 for RAO 3 are generally

lower than RALs (but in some cases, RAO 3 RALs are the same as the cleanup levels). Cleanup levels are based on

state or federal standards (whichever value is more stringent) and if no standard exists then risk-based

concentrations are developed. At this site cleanup levels for sediment are based on Sediment Cleanup Objectives

(SCOs) from the State Sediment Management Standards (SMS). See text box on page 26 for more information

about the SMS. These levels must to be achieved post-construction, or after a period of monitored natural recovery

(MNR). Achievement of cleanup levels for RAOs 1, 2, and 4 is measured by averaging sample results over specific

areas using the UCL95 value (see Table 19). Achievement of cleanup levels for RAO 3 are measured at each

sampling station (see Table 20).


124

SCO after dredging to sufficient depth to accommodate a cap, all contaminated sediments will be

dredged. If greater than 1 ft of contamination would remain after dredging to sufficient depth to

accommodate a cap, sediments will be partially dredged and capped.

All post-remedy surfaces within the federal navigation channel will be maintained at or below

their current authorized depths. In order to avoid damage to a cap or ENR layer during federal

maintenance dredging, the top of any ENR layer will be at least 2 ft and the top of any cap will be

at least 4 ft below the authorized federal navigation channel depth. For areas outside the

navigation channel where depths are maintained by private or public entities (called berthing

areas in this ROD, but could include slips, entrance channels, or restorations areas) the top of any

cap or ENR layer will be a minimum of 2 ft below the operating depth.

In habitat areas24

, post-remedy surfaces will be maintained at their current depth and backfilled or

capped with suitable habitat materials.

Dredging may be required in some areas that would otherwise be designated for capping if ICs

required to prevent damage to a cap (such as prohibitions on tug maneuvering or use of spuds

[vessel-mounted poles that are sunk into sediment for stabilizing vessels]) are not compatible

with the current or reasonably anticipated future use of that area. See Sections 13.2.3 and 13.2.4.

An additional 10 ft (lateral) of dredging outside of the federal navigation channel will be included

to assure that side slopes are stable and do not slough into the channel.

Dredged materials will be transported via truck or rail for disposal at a permitted upland off-site landfill

facility.25

Engineered sediment caps will be placed in areas where sediments are more highly contaminated (COC

concentrations are higher than ENR upper limits; see Section 13.2.1.2, and Table 28) where there is

sufficient water depth for a cap. Caps in intertidal clamming areas must include a minimum 45 cm clam

habitat layer. EPA estimates that caps in intertidal clam habitat areas will generally be 4 ft thick. In other

areas, cap thickness will generally be 3 ft. Cap thickness will be evaluated during remedial design in

accordance with EPA and USACE (1998). In habitat areas, the uppermost layers of caps will be designed

using suitable habitat materials. Other materials, such as activated carbon or other contaminant-

sequestering agents, may be used to reduce the potential for contaminants to migrate through the cap.

24

For the FS, all areas above -10 ft MLLW were assumed to be habitat areas for the purpose of developing remedial

alternatives. As part of the remedial design, EPA, in coordination with natural resource agencies and Tribes, will

determine what areas are considered habitat areas for the purpose of complying with ESA and Section 404 of the

CWA (see Table 26). EPA will also determine what elevations and what substrate materials will be required for

caps, ENR, or placement of backfill materials in any identified habitat area. 25

Some clean materials may be dredged as part of the cleanup; for example, in order to maintain appropriate

sideslopes at the edge of a dredge cut. Sediments that pass the Dredged Materials Management Program’s

criteria may be disposed at an open-water disposal site.


125

Table 28. Remedial Action Levels, ENR Upper Limits, and Areas and Depths of Application

Intertidal Sediments (+11.3 ft MLLW to -4 ft MLLW) Subtidal Sediments (-4 ft MLLW and Deeper)

Recovery Category 1 RALs, ENR ULs, and Application Depths

Recovery Category 2 and 3 RALs, ENR ULs, and Application Depths

Recovery Category 1 RALs, ENR ULs, and Application Depths

Recovery Category 2 and 3 RALs, ENR ULs, and Application Depths

Shoaled Areasb in Federal Navigation Channel

Risk Driver COC Units

Action Levels Top 10 cm (4 in) Top 45 cm (1.5 ft) Top 10 cm (4 in) Top 45 cm (1.5 ft) Top 10 cm (4 in) Top 60 cm (2 ft) Top 10 cm (4 in) Top 60 cm (2 ft)c

Top to Authorized Navigation Depth Plus 2 ft

Human Health Based RALs

PCBs (Total) mg/kg OC RAL 12 12 12 65 12 12 12 195 12

ULa for ENR -- -- 36 97 -- -- 36 195 --

Arsenic (Total) mg/kg dw RAL 57 28 57 28 57 57 57 -- 57

ULa for ENR -- -- 171 42 -- -- 171 -- --

cPAH µg TEQ/kg dw RAL 1000 900 1000 900 1000 1000 1000 -- 1000

ULa for ENR -- -- 3000 1350 -- -- 3000 -- --

Dioxins/Furans ng TEQ/kg dw RAL 25 28 25 28 25 25 25 -- 25

ULa for ENR -- -- 75 42 -- -- 75 -- --

Benthic Protection RALs

39 SMS

COCs d

Contaminant-specific

RAL Benthic SCO Benthic SCO 2x Benthic SCO -- Benthic SCO Benthic SCO 2x Benthic SCO -- Benthic SCO

ULa for ENR -- -- 3x RAL -- -- -- 3x RAL -- --

a. The ENR Upper Limit (UL) is the highest concentration that would allow for application of ENR in the areas described. For areas with no ENR limit listed, ENR is not a currently designated technology (see Section 13.2.1.2 for further discussion).

b. Shoaled areas are those areas in federal navigation channel with sediment accumulation above the authorized depth including a 2 ft over-dredge depth that USACE uses to maintain the channel for navigation purposes. The authorized channel depths are (1) from RM 0 to 2 (from Harbor Island to the First Avenue South Bridge), 30 ft below MLLW; (2) from RM 2 to RM 2.8 (from the First Avenue South Bridge to Slip 4), 20 ft below MLLW; and (3) from RM 2.8 to 4.7 (Slip 4 to the Upper Turning Basin), 15 ft below MLLW. For shoaled areas, the compliance intervals will be determined during Remedial Design; these are typically 2-4 ft core intervals. For areas in the channel that are not shoaled, Recovery Categories 1 or 2 & 3 RALs apply as indicated in the other subtidal columns.

c. Applied only in potential vessel scour areas. These are defined as subtidal areas (i.e., below -4 ft MLLW) that are above -24 ft MLLW north of the 1st Ave South Bridge, and above -18 ft MLLW south of the 1st Ave South Bridge (see Figure 17).

d. There are 41 SMS COCs, but total PCBs and arsenic ENR ULs are based upon human health based RALs only (see Table 20).


126

[This page left blank for double-sided printing.]


127

In Recovery Category 1 areas, dredging, capping, or a combination thereof is required when any of

the conditions listed below have been met:

In intertidal and subtidal areas, in Recovery Category 1, any sediment COC concentration

averaged over the top 10 cm is greater than any of the benthic protection RALs (benthic SCO

criteria, see Table 27) or greater than any of the four human health RALs (PCBs, arsenic, cPAHs,

dioxins/furans, see Table 28 and Figures 22 and 23).

In intertidal areas in Recovery Category 1, sediment COC concentrations averaged over the top

45 cm are greater than any of the four human health RALs.

In subtidal areas in Recovery Category 1, sediment COC concentrations averaged over the top 60

cm are greater than any of the benthic protection RALs or greater than any of the four human

health RALs.

In Recovery Category 2 and 3 areas, dredging, capping, or a combination thereof is required when

COC concentrations exceed the criteria for application of ENR described in Section 13.2.1.2. See Figure

19 and Figure 20.

13.2.1.2 ENR

A thin layer (6 to 9 inches) of clean material will be placed (referred to as enhanced natural recovery

[ENR]) in areas that meet the criteria for ENR as described below. Suitable habitat materials will be used

in habitat areas. ENR may include in situ treatment using activated carbon or other amendments, and

engineered designs for sediment stability. The effectiveness and potential impacts of using in situ

treatment or amendment technologies, as well as the areas best suited for these technologies, will be

evaluated in pilot studies performed during remedial design.

In Recovery Category 2 and 3 areas, ENR with or without in situ treatment will be selected based on

sediment COC concentrations and the potential for sediment scour (Table 28 and Figure 17):

In intertidal areas in Recovery Categories 2 and 3, ENR will be applied when any sediment COC

concentration averaged over the top 10 cm is between 1 and 3 times the top 10 cm intertidal

RALs (e.g., 12 – 36 mg/kg OC PCBs), or when any sediment COC concentration averaged over

the top 45 cm is between 1 and 1.5 times the intertidal RALs for the 45 cm interval (e.g., 65 – 97

mg/kg OC PCBs).

In subtidal areas in Recovery Categories 2 and 3, ENR will be applied when any sediment COC

concentration in the top 10 cm is between 1 and 3 times the top 10 cm subtidal RALs. In potential

vessel scour areas26

(Figure 17), sediment concentrations of PCBs averaged over the top 60 cm

must also be less than 3 times the CSL chemical criterion (195 mg/kg OC). There are no RALs

for the top 60 cm in Category 2 and 3 areas in deeper water depths; in these areas, RALs are

applied only to the top 10 cm.

Pilot testing will be performed to determine whether ENR/in situ treatment is effective in

reducing toxicity and bioavailability of COCs while avoiding unacceptable impacts to biota. If

pilot testing shows that ENR/in situ treatment can meet these objectives, EPA will consider, in

coordination with the state and Tribes, the locations where ENR with in situ treatment will be

applied. These areas may include some of the Recovery Category 1 areas where it can be

26.

Subtidal areas in Recovery Categories 2 and 3 deemed to be potentially subject to vessel scour especially by

tugboats are: north of the 1st Avenue South Bridge (located at approximately RM 2) in water depths from -4 to -

24 ft MLLW, and south of the 1st Avenue South Bridge, in water depths from -4 to -18 ft MLLW. These depths

are based on the size of tugboats that normally operate in these areas.


128

demonstrated that ENR with in situ treatment will maintain its stability and effectiveness in these

areas over time; for example, areas where vessel- and flood-related scour were shown by the

STM and FS scour analysis to be minor. EPA may also consider ENR with in-situ treatment in

areas with COC concentrations up to the CSL if it can be demonstrated that it will maintain its

effectiveness over time.

ENR will not be applied to Recovery Category 1 areas unless EPA approves it, as discussed

above.

13.2.1.3 Other Considerations for Application of Cleanup Technologies

EPA will apply location-specific cleanup technologies to areas with structural or access restrictions (e.g.,

under-pier areas and in the vicinity of dolphins/pilings, bulkheads, and riprapped or engineered

shorelines). Debris and pilings will be removed throughout the LDW as necessary or as required by EPA

to implement the remedy, and materials will be disposed at a permitted off-site facility.

13.2.2 Monitored Natural Recovery

MNR will be applied in all areas of the LDW that are not remediated through capping, dredging, or ENR.

For all areas where MNR is applied, long-term monitoring of surface sediments (top 10 cm) will be

implemented to evaluate whether the RAO 3 cleanup levels (benthic SCO criteria) are being achieved in a

reasonable timeframe or are not met within 10 years after remediation. The STM and BCM, supported by

data collected during the RI/FS, were used to estimate the amount of time required to reduce COC

concentrations in sediments through natural recovery. The STM and BCM natural recovery predictions

will be reevaluated using data collected during remedial design.

MNR To Benthic SCO will be applied where the concentration of any of the 39 RAO 3 COCs

(i.e., excluding the human health COCs PCBs and arsenic) is less than the RAL but greater than

the RAO 3 cleanup levels (benthic SCO criteria; Table 27 and Figure 21), and modeling results

indicate the COC will be reduced to the benthic SCO criteria within 10 years of the completion of

remedial action. More intensive long-term monitoring will be conducted in these areas, and

should MNR not achieve RAO 3 cleanup levels or progress sufficiently toward achieving them in

10 years, additional actions (dredging, capping, or ENR) will be implemented. Those actions will

be determined using the same approach set forth in this decision document as described in

Section 13.2.1 and illustrated in Figures 19 and 20.

MNR Below Benthic SCO will be applied where the concentration of all COCs is less than the

RAL and the RAO 3 cleanup levels (benthic SCO criteria), but greater than the human health-

based (RAO 1 and 2) cleanup levels (which are measured on an LDW-wide or area-wide basis,

see Table 19 and Figure 21). Less intensive monitoring will be conducted in these areas. If

cleanup levels are not achieved, additional cleanup actions may be considered and selected in a

future decision document, see Section 13.4.

13.2.3 Monitoring The entire LDW will be sampled as part of baseline, construction, post-construction, and long-term

monitoring.

Remedial design sampling and analysis will be conducted to establish post-EAA cleanup

baseline conditions. Remedial design sampling data will be used to refine the cleanup footprint

shown in Figure 18 using the decision criteria described in Figure 19 through Figure 22. Results

will also be used to evaluate of the effectiveness of EAA cleanups and the degree to which

natural recovery has occurred since the RI/FS sampling, to serve as a baseline for comparison to


129

post-cleanup data, and to aid in the evaluation of source control effectiveness. Remedial design

sampling will include:

o Establishing baseline contaminant concentrations in surface and subsurface sediments,

surface water, and porewater. Sediment samples will be analyzed for all RAO 1, 2, 3, and

4 COCs (Table 19, Table 20, and Table 21); and a subset of sediment samples will be

analyzed for other contaminants not selected as COCs but identified in the HHRA as

posing an excess cancer risk of greater than 1 x 10-6

or noncancer HQ of 1 at the adult

Tribal RME consumption rate (see Table 14), to assess their reduction over time, as well

as to determine conventional and engineering parameters. Biological testing (benthic

community toxicity and abundance) will be included as determined during remedial

design. Surface water samples will be initially analyzed for all analytes in Washington

WQS (WAC173-201A), AWQC (CWA Section 304[a]) and NTR (40 CFR 131.36(b)(1)

as applied to Washington, 40 CFR 131.36(d)(14)). Following the first few sampling

rounds, the surface water analyte list will be reduced to the contaminants that exceeded

AWQC, NTR, or Washington WQS values.

o Sampling to better understand the concentrations of incoming suspended sediments from

the Green/Duwamish River that deposit in the LDW, in order to refine the RI/FS BCM

predictions and inform the long-term monitoring program.

o Measuring contaminant concentrations in fish and shellfish tissue in the LDW to inform

fish advisories and to provide a baseline to measure the success of the remedial action in

reducing fish and shellfish tissue concentrations (RAO 1). Samples will be analyzed for

PCBs, arsenic, cPAHs, and dioxins/furans; and a subset of tissue samples will be

analyzed for other contaminants not selected as COCs but identified in the HHRA as

posing an excess cancer risk of greater than 1 x 10-6

or noncancer HQ of 1 at the adult

Tribal RME consumption rates (see Table 14). Additional fish and shellfish tissue data

will also be collected in non-urban areas in Puget Sound to refine the non-urban

background values (see Table 4) that will be used for comparison to Site data to measure

progress in reducing tissue concentrations.

o Conducting research to further assess the relationship between arsenic and cPAH

concentrations in sediment and in clam tissue, and to assess whether remedial action can

reduce clam tissue concentrations to achieve RAO 1. EPA anticipates that

implementation of the Selected Remedy, along with implementation of source control

actions, will achieve the RAO 2 (direct contact) cleanup levels for arsenic and cPAHs,

which will also result in lower clam inorganic arsenic and cPAH concentrations that will

achieve RAO 1 the extent practicable; however, at this time, the amount of reduction is

uncertain. If EPA determines, based on these studies, that additional remedial action is

needed to reduce clam tissue arsenic and cPAH concentrations for the purpose of

achieving RAO 1, EPA will document and select those actions in a future decision

document.

Recovery Category areas will be re-evaluated during remedial design. The criteria for

Recovery Categories (Table 23) were applied in the FS (LDWG 2012a) based upon best available

knowledge using best professional judgment. EPA will use additional information and analysis

and the criteria in Table 23 to change Recovery Category assignments in specific areas of the

LDW where appropriate. Information EPA will consider in deciding whether to modify recovery

categories include the following:


130

o A survey of waterway users, including tribal members exercising their treaty rights, will

be conducted to gather detailed information about waterway use, including tribal fishing;

maneuvering and anchoring of ships, barges and tugs; use of spuds; and other activities

such as berth and wharf maintenance. Information about such activities may change

Recovery Categories of some areas.

o EPA will also consider other information such as refined sedimentation rates and

contaminant trends based upon new data. EPA will also reconsider areas where the

Recovery Category designation in the FS appears to have deviated from the criteria in

Table 23.

Monitoring during and after construction will include environmental monitoring to ensure

compliance with RALs and ARARs, and monitoring of physical as-built conditions (e.g.,

bathymetry) to ensure compliance with construction standards and project design documents.

Long-term monitoring of sediments, surface water, porewater, fish and shellfish tissue and

benthic community toxicity and abundance will be conducted to ensure protectiveness of human

health and the environment, to ascertain attainment of cleanup levels and compliance with

ARARs, to protect the integrity of the remedial actions, and to aid in the evaluation of source

control effectiveness.

If any habitat areas are constructed as part of the remedial action to comply with CWA Section

404, baseline and long-term monitoring will include appropriate habitat monitoring.

The details of long-term monitoring and maintenance, including performance standards, sampling

density and frequency, interim benchmarks, and associated additional actions, as well as maintenance

of remedy elements such as caps, ENR areas, and habitat areas, will be provided in a long-term

monitoring and maintenance plan to be developed in remedial design. Samples will be analyzed for

the analytes listed above for baseline sampling, with the list modified during remedial design based

on baseline results.

13.2.4 Institutional Controls

Institutional controls will be required for the entire waterway to reduce human exposure to contaminants,

and protect the integrity of the remedy. However, reliance on ICs, particularly seafood consumption-

related ICs, will be minimized to the extent practicable. ICs include proprietary controls in the form of

Washington Uniform Environmental Covenants Act (UECA)-compliant environmental covenants, and

informational devices including fish and shellfish consumption advisories to reduce human exposure from

ingestion of contaminated resident fish and shellfish. EPA anticipates relying on the existing WDOH fish

and shellfish consumption advisories (see Section 6.2), and information obtained through the ongoing

study of fishing and fish and shellfish consumption patterns (Fishers Study: LDWG 2014b) will be used

to develop appropriate and effective ICs, which will include other measures to provide additional

protectiveness, such as outreach and education programs.

As noted in Section 13.2.3, EPA will gather detailed information in remedial design about waterway use

in specific areas, including impacts on tribal treaty rights. EPA will use that information to develop

location-specific use restrictions (environmental covenants or governmental controls, such as restricted

navigation areas designated by the Coast Guard) that would prohibit activities that may damage caps such

as tug maneuvering and spudding. If such ICs interfere with waterway activities required for use of a

particular area, dredging may be required instead of capping to allow for fewer restrictions on the use of

the area.


131

13.2.5 Use of Green Remediation Practices

To the extent practicable, the remedial action should be carried out consistent with EPA Region 10’s

Clean and Green policy (EPA 2009b), including the following practices:

Use renewable energy and energy conservation and efficiency approaches, including Energy Star

equipment.

Use cleaner fuels such as low-sulfur fuel or biodiesel, diesel emissions controls and retrofits, and

emission reduction strategies.

Use water conservation and efficiency approaches including Water Sense products.

Use reused or recycled materials within regulatory requirements.

Minimize transportation of materials and use rail rather than truck transport to the extent

practicable.

13.2.6 Role of EAAs in the Selected Remedy

Dredging, capping, ENR, and MNR as described above apply to 412 acres of the LDW. An additional 29

acres of the most contaminated sediments in the LDW have been or will be addressed by cleanups in

Early Action Areas (described in Sections 2.3 and 4.1). EPA has reviewed the EAA cleanup actions

subject to implementation under EPA Consent Orders (for the Slip 4, Terminal 117, Boeing Plant 2, and

Jorgensen Forge facilities), and has determined that the completed Slip 4 EAA is consistent with the

Selected Remedy and requires no further active remediation. The other planned EAA cleanups conducted

under EPA oversight are similarly expected to require no further active remediation if they achieve their

stated objectives. For the cleanups conducted under the 1991 Natural Resource Damages Consent Decree

(Norfolk CSO/SD and Duwamish/Diagonal CSO/SD), EPA will conduct a review during the remedial

design phase to determine whether additional work is needed to make these cleanup actions consistent

with the remedy selected in the ROD. EPA will review the IC plans and long-term monitoring plans for

all of the EAAs and will require that the EAAs be incorporated into plans for the rest of the LDW as

necessary to make them consistent with the Selected Remedy.

13.2.7 Role of Source Control in the Selected Remedy

The Selected Remedy will be implemented while a comprehensive source control program is managed by

Ecology, as described in the Source Control Strategy, which will be updated after completion of the ROD.

EPA and Ecology will coordinate before initiating active in-waterway cleanup to ensure that sources have

been sufficiently controlled to prevent or minimize the likelihood that sediment will be recontaminated

before initiating active remediation in any portion of the waterway (see Section 4.2). The coordination

process is further explained in a 2014 Memorandum of Agreement (MOA), and in Ecology's Source

Control Strategy. EPA’s draft Implementation Plan, which was provided to Ecology in 2013, provides

additional details on the coordination process among EPA offices and with Ecology.

This ROD addresses the In-waterway Portion of the Site only and does not impose requirements on or in

any way limit Ecology in its implementation of source control under State law, including MTCA and the

WPCA. Furthermore, this ROD does not limit Ecology's implementation of Clean Water Act delegated

authorities. Over time, the integrated approach of CERCLA and longer-term clean water actions are

expected to result in attainment of applicable surface water criteria and uses under the Clean Water Act.


132

13.2.8 Addressing Environmental Justice concerns

Environmental Justice concerns will be addressed before, during, and after implementation of the remedy

through means that include the following:

Reducing human health risks as quickly as practicable, while also providing for long-term

effectiveness and permanence.

Conducting the Fishers Study (LDWG 2014b) to learn more about the affected community (those

who consume LDW resident fish and shellfish) in order to enhance outreach efforts. As noted in

Section 10.3.3, EPA has already started implementing this recommendation as part of the RI/FS.

Continuing to engage the community throughout remedial design and implementation of the

cleanup, including convening an advisory group as a means for the affected community and local

agencies to work together on mitigating the impacts of the cleanup on the affected community.

Continuing consultation with affected Tribes on recommendations for the remedy.

Reducing the impacts of the cleanup on residents through green remediation techniques, as

discussed in Section 13.2.5.

13.3 Cost Estimate for the Selected Remedy

The information presented in the cost estimate summary table for the Selected Remedy is based on the

best available information regarding its anticipated scope. Changes in the cost elements are likely to occur

as a result of the new information and data collected during remedial design. Major changes may be

documented in the form of a memorandum to the Administrative Record file, an ESD, or a ROD

amendment. This is an order-of-magnitude engineering cost estimate that is expected to be within +50 to

-30 percent of the actual project cost. Table 25 compares costs for all alternatives and the Selected

Remedy, using 0%, 2.3%, and 7% discount rates. Table 29 presents a detailed cost estimate for the

Selected Remedy at the 2.3% discount rate.


133

Table 29. Cost Estimate Summary for Selected Remedy

ELEMENT UNIT COSTS UNIT QUANTITY /

SUBTOTAL

PRECONSTRUCTION

Mob, Demob & Site Restoration (project) $ 800,000 Lump Sum 1

Mob, Demob & Site Restoration (seasonal) $ 120,000 YEAR 10.5

Land Lease for Operations & Staging $ 250,000 YEAR 10.5

Contractor Work Plan Submittals $ 100,000 YEAR 10.5

Barge Protection $ 80,000 Lump Sum 1

Subtotal: $ 5,813,932

PROJECT MANAGEMENT (CONTRACTOR)

Labor & Supervision $ 62,000 MONTH 48.3

Construction Office & Operating Expense $ 21,600 MONTH 48.3

Subtotal: $ 4,037,006

DREDGING

Shift Rate $ 25,963 DAY 924

Gravity Dewatering (on the barge) $ 10 CY 950,664

Subtotal: $ 33,496,452

SEDIMENT HANDLING & DISPOSAL

Transloading Area Setup $ 1,000,000 Lump Sum 1

Water Management $ 10,000 DAY 924

Transload, Railcar Transport to & Tipping at Subtitle D Landfill $ 60 TON 1,425,997

Subtotal: $ 95,799,820

SEDIMENT CAPPING, DREDGE RESIDUALS, DREDGE BACKFILL

Debris Sweep $ 30,000 ACRE 2

Shift Rate (12 hours) $ 12,500 DAY 501

Cap Material Procurement & Delivery (sand) $ 27 CY 548,103

Subtotal: $ 21,121,281

ENHANCED NATURAL RECOVERY

Debris Sweep $ 30,000 ACRE 5

Shift Rate (12 hours) $ 12,500 DAY 46

Material Procurement & Delivery (sand) $ 27 CY 28,824

Material Procurement & Delivery (carbon amended sand) $ 161 CY 28,824

Subtotal: $ 6,143,912

CONSTRUCTION QA/QC

Construction Monitoring $ 7,925 DAY 924

Subtotal: $ 7,322,700

POST-CONSTRUCTION PERFORMANCE MONITORING

Compliance Testing (Dredging)

PROJECT $ 1,445,267

Compliance Testing (Capping)

PROJECT $ 1,141,320

Compliance Testing (ENR)

PROJECT $ 1,221,569

Subtotal: $ 3,808,157

CAPITAL COSTS (base) $ 177,543,260

CAPITAL COSTS (present value) $ 159,745,069


134

ELEMENT UNIT COSTS UNIT QUANTITY /

SUBTOTAL

Construction Contingency 35% PROJECT $ 62,140,141

Sales Tax 9.5% PROJECT $ 16,866,610

Project Management, Remedial Design & Baseline Monitoring 30% PROJECT $ 53,262,978

Construction Management 10% PROJECT $ 17,754,326

TOTAL CAPITAL COST (base) $ 327,567,314

TOTAL CAPITAL COST (present value)

$ 294,729,653

AGENCY OVERSIGHT, REPORTING, O&M, & MONITORING COSTS (base)

Agency Review & Oversight

PROJECT $ 10,200,000

Reporting

PROJECT $ 1,900,000

Operations & Maintenance (Dredging)

PROJECT $ 1,416,056

Operations & Maintenance (Capping)

PROJECT $ 5,907,000

Operations & Maintenance (ENR)

PROJECT $ 6,352,496

Operations & Maintenance (MNR>SCO)

PROJECT $ 2,250,956

Operations & Maintenance (MNR<SCO)

PROJECT $ 8,978,076

Long-term Monitoring

PROJECT $ 5,775,580

Institutional Controls

PROJECT $ 25,000,000

Subtotal (base): $ 67,780,164

Subtotal (present value):

$ 47,504,279

TOTAL COST (Net Present Value) at 2.3% discount rate $ 342,233,932

13.4 Estimated Outcomes of Selected Remedy

The intent of the Selected Remedy is, in conjunction with cleanup of the EAAs and with Ecology-led

source control activities, to be protective of human health and the environment and to attain ARARs,

although some ARARs may not be achieved for the foreseeable future. It is consistent with current and

reasonably anticipated future uses of the waterway. It is intended to minimize reliance on fish and

shellfish consumption-related institutional controls to the extent practicable; however, such controls will

have to remain in effect to ensure protectiveness for the foreseeable future.

The goal of this CERCLA cleanup action and the Ecology-led source control program is to reduce in-

waterway contamination and sources to the waterway to levels needed to achieve all cleanup levels and

ARARs described in Section 8 and Table 19 and Table 20. RI/FS modeling results conclude that it may

not be possible for any alternative to do so; however, as discussed in Sections 8 and 10, it is difficult to

predict long-term Site conditions with any degree of accuracy.

The active remedy components of the Selected Remedy are expected to take 7 years to implement after

completion of the EAAs and remedial design, and after sources have been sufficiently controlled to

minimize recontamination (see Section 4.2). The Selected Remedy will be designed to maintain sufficient

water depth for human use and habitat function and allow for future navigation dredging. During and

after remediation current and anticipated future land and waterway uses, including industrial, residential,

commercial and recreational uses, are expected to be able to continue, subject to the institutional controls

and so long as sources of contamination are controlled or eliminated. EPA expects that direct contact risks

(RAO 2) and risks to higher trophic level species (RAO 4) will be reduced to the cleanup levels (except as


135

noted in Section 10.1.2) and risks to benthic invertebrates (RAO 3) and human seafood consumers (RAO

1) will have been significantly reduced at the completion of active components of the remedy. EPA

anticipates that another 10 years of natural recovery will be required to reduce COC concentrations

sufficiently to meet RAO 3 and RAO 1 to the extent practicable.

The lowest contaminant concentrations in fish and shellfish tissue are predicted by modeling to be

achieved in 17 years following the start of construction. EPA will review long-term monitoring data to

assess the success of the remedy, including measuring contaminant concentrations in sediment, surface

water, and fish and shellfish tissue. If long-term monitoring data show that RAO 3 cleanup levels (benthic

SCO criteria) and human health-based RALs (see Table 27 and Table 28) are exceeded, additional actions

will be taken to reduce COC concentrations to these levels. If monitoring shows that contaminant

concentrations have reached a steady state at levels below the benthic SCO criteria or human health-based

RALs but above the human health risk reduction or background-based cleanup levels, EPA will review

the data and consider whether additional technically practicable cleanup actions would further reduce

contaminant concentrations in sediments, tissue, or surface water.

EPA expects that, once the active components of the Selected Remedy (dredging, capping, ENR, and any

additional actions needed to meet the benthic SCO criteria and human health-based RALs) have been

completed and long-term monitoring shows COC concentrations have reached a steady state, COC

concentrations will either be at cleanup levels for sediment and ARARs for water quality, or will

represent practicable limitations in implementation of source control and active remediation. Data

collection and analysis during long-term monitoring is intended to test this expectation.

However, if EPA determines that additional remedial action is appropriate for the In-waterway Portion of

the Site, EPA will select such action in a ROD Amendment or ESD. If EPA or the State determines that

further source control is appropriate, EPA or the State will address such sources with source control

response action decisions separate from this ROD. If EPA determines that no additional practicable

actions can be implemented under CERCLA to meet ARARs, EPA may issue a ROD Amendment or

ESD providing the basis for a technical impracticability waiver for specified sediment and/or surface

water quality based ARARs under Section 121(d)(4)(C) of CERCLA.

Implementation of the Selected Remedy, along with the EAA cleanups and source control, will

substantially improve the quality of LDW sediments and surface water, reduce COC concentrations in

waterway organisms, and result in an estimated 90% or greater reduction in seafood consumption risk. It

should also address the key Environmental Justice concerns as discussed in Section 13.2.8.


136

Figure 17. Recovery Category 1 and Potential Tug Scour Areas in LDW


137

Figure 18. Selected Remedy


138



139

Figure 19. Intertidal Areas – Remedial Technology Applications


140



141

Figure 20. Subtidal Areas – Remedial Technology Application


142



143

Figure 21. Intertidal and Subtidal Areas – Natural Recovery Application


144

Figure 22. Intertidal Areas - Remedial Action Levels Application


145

Figure 23. Subtidal Areas – Remedial Action Levels Application


146



147

14 Statutory Determinations

CERCLA Section 121 and the NCP Section 300.430(f)(5)(ii) require selection of a remedy or remedies

that are protective of human health and the environment, comply with ARARs (unless a statutory waiver

is justified), are cost-effective, and use permanent solutions and alternative treatment technologies to the

maximum extent practicable. In addition, CERCLA includes a preference for remedies that employ

treatment that permanently and significantly reduces the volume, toxicity, or mobility of hazardous

substances, pollutants, or contaminants as a principal element.

As discussed below, EPA has determined that the Selected Remedy meets these statutory requirements.

14.1 Protection of Human Health and the Environment

The Selected Remedy will protect human health and the environment through dredging or capping the

most contaminated sediments, using ENR to reduce COC concentrations in moderately contaminated

areas, and using MNR to further reduce concentrations in less contaminated areas.

Long-term model projections of risk reduction achieved by the Selected Remedy 30 - 40 years in the

future are uncertain. The FS estimated an adult Tribal RME excess cancer risk of 2 x 10-4

and noncancer

HQ of 4 and a child Tribal RME excess cancer risk of 5 x 10-5

and noncancer HQ of 9 for PCBs at the

model-predicted steady state after implementation of the cleanup. As discussed in Section 8.2.1, model

projections in the FS may underestimate the long-term risk reduction achieved by the Selected Remedy. If

the target tissue concentrations in Table 21 are achieved, excess cancer risks for PCBs are estimated to be

5 x 10-6

for the adult Tribal RME scenario and 1 x 10-6

for the child Tribal RME scenario and noncancer

HQs are estimated to be less than 1 for both adult and child Tribal RME scenarios. Compared to the adult

and child Tribal RME seafood consumption rate baseline risks presented in Section 7.1, these estimates of

post-cleanup risks represent a reduction in PCB risks of approximately 90% at the model-predicted steady

state and 99% at the target tissue concentration.

Long-term projections of risk at the model-predicted steady state could only be made for PCBs. Estimated

cancer and noncancer risks for the adult Tribal RME seafood consumption rate if fish and shellfish target

tissue concentrations are achieved for all COCs (which, for some COCs, are based on natural background

levels that are higher than calculated protective risk-based levels (RBTCs-see Table 21) are estimated to

be 3 x 10-4

, and the HQ for noncancer risks would be less than 1 (based on PCBs, the COC with the

highest HQ).

As discussed in Section 4.3, and other places in this document, institutional controls limiting seafood

consumption will be needed for the foreseeable future. The intent of the remedy is to reduce contaminant

concentrations in sediments, surface water, and fish and shellfish tissue to the extent practicable, and to

minimize reliance on seafood consumption advisories to attain protectiveness.

14.2 Compliance with Applicable or Relevant and Appropriate Requirements

ARARs for the Selected Remedy are shown in Table 26. The most significant ARARs for this remedial

action are the MTCA/SMS requirements and Federal and State water quality criteria and standards, as

discussed in Section 8.2.2. The objective of the Selected Remedy is to meet ARARs throughout the In-

waterway Portion of the Site.


148

MTCA/SMS

The sediment cleanup levels in this ROD are based on the SCO criteria in the SMS. The SCO-based

sediment cleanup levels for protection of human health are set at natural background concentrations or at

the RBTCs27

, whichever is higher. The SMS allows for selection of higher CSL-based cleanup levels

when it is not technically possible to achieve SCO levels, or if meeting the SCO will have a net adverse

impact on the aquatic environment (WAC 173-204-560). EPA has determined that there is insufficient

information at this time to determine whether it is technically possible to achieve the SCO-based cleanup

levels selected in this ROD, as described in Section 8.2.2.1. The selected cleanup levels for the protection

of benthic invertebrates (RAO 3) are the SCO chemical and biological criteria set forth in WAC 173-204-

562. The SMS also requires protection of HTLS; cleanup levels for the protection of human health and

benthic invertebrates are also protective of HTLS.

The Selected Remedy is predicted to reduce contaminant concentrations to the RAO 3 cleanup levels (the

benthic SCO) within 10 years after the completion of active cleanup. Additional actions (see Section

13.2.2) will be implemented if they are not achieved within 10 years. However, as discussed in Section

10.1.1, the RI/FS models indicate that the long-term COC sediment concentrations achievable in the In-

waterway Portion of the Site will be limited by the extent to which all ongoing sources, including COCs

entering the waterway from the upstream Green/Duwamish River system and remaining lateral sources,

can be controlled in this urban environment. Specifically, the RI/FS models indicate that while

remediation is predicted to result in significant improvements in sediment and tissue COC concentrations,

implementation of the Selected Remedy is not predicted to achieve the SCO-based sediment cleanup

levels required by the SMS for PCBs, dioxins/furans (for RAO 1), and arsenic (for RAO 2).

As discussed in Sections 8.2.1 and 8.2.2, the Selected Remedy could potentially meet ARARs

notwithstanding these RI/FS model projections. See Section 8.2.1 for a discussion of RI/FS modeling

uncertainties. The objective of the Selected Remedy and Ecology's source control program under its

authorities is to reduce COC concentrations to cleanup levels protective of human health, HTLS, and

benthic invertebrates, as required by the SMS and to meet water quality ARARs. EPA will monitor

progress of these actions, including measuring long-term contaminant concentrations in sediment and

surface water for ARAR compliance. EPA will also monitor fish and shellfish tissue concentrations to

inform decision making with respect to protectiveness of human health and the environment, including

the need for and content of institutional controls. Institutional controls are required by MTCA, WAC 173-

340-440(a)(4), when hazardous substances above cleanup levels remain at a site, which is wholly

consistent with EPA policy and guidance. This MTCA rule is an ARAR.

If long-term monitoring data and trends indicate that some ARARs cannot be met, EPA will determine

whether further In-waterway remedial action in conjunction with source control could practicably achieve

the ARAR. If EPA concludes that an ARAR cannot be practicably achieved, EPA will either waive the

ARAR on the basis of technical impracticability (TI) in a future decision document (ROD Amendment or

ESD), or for SMS SCO-based ARARs, EPA will consider whether the criteria in the SMS for adjusting

cleanup levels upward from the SCO, to no higher than the CSL, can be met as discussed above. If these

27. MTCA/SMS require a total excess cancer risk of less than or equal to 1 x 10

-5 and

excess cancer risks for

individual COCs less than or equal to 1 x 10-6

and a noncancer HQ or HI of less than 1 for protection of human

health.


149

criteria can be met, EPA will evaluate adjusting the relevant sediment cleanup levels upward to regional

background or other CSL-based levels described in the SMS.

Because EPA cannot know whether or to what extent the SMS ARARs for various COCs will be

achieved upon completion of remedial action (including natural recovery), consideration of such

waiver(s) prior to the collection of monitoring data sufficient to inform TI waiver decisions, or upward

adjustment of cleanup levels under the SMS, is neither warranted nor justifiable at this time.

Surface Water Quality ARARs

Surface water quality ARARs consist of applicable promulgated state water quality standards and, in

accordance with Section 121(d)(2)(A)(ii) and (B)(i) of CERCLA, federal recommended Clean Water Act

Section 304(a) Ambient Water Quality Criteria (AWQC) guidance values where they are relevant and

appropriate, as discussed in Section 8.2.2.2. Current concentrations of PCBs in the upstream

Green/Duwamish River surface water are higher than the selected ARARs (discussed above and in

Section 8.2.2.2). RI/FS model projections assumed no future decrease in the current upstream surface

water COC concentrations. Implementation of the sediment remedy in combination with source control

implementation under State-lead authority will improve surface water quality to an unknown degree. If

appropriate, waiver of surface water quality ARARs will be considered only after the improvement from

these combined actions is assessed based on long-term water quality monitoring. As discussed above, if

long-term monitoring indicates that surface water quality ARARs cannot be met, EPA will review the

data and consider whether additional technically practicable cleanup would further reduce contaminant

concentrations in surface water. If EPA determines that additional remedial action is appropriate for the

In-waterway Portion of the Site, EPA will select such action in a ROD Amendment or ESD. If EPA or the

State determine that further source control is appropriate, EPA or the State will address such sources with

source control response action decisions separate from this ROD. If EPA determines that no additional

practicable actions can be implemented to meet ARARs, EPA may issue a ROD Amendment or ESD

providing the basis for a technical impracticability waiver for water-quality based ARARs under Section

121(d)(4)(C) of CERCLA.

ESA

To protect threatened or endangered species under the ESA, including Puget Sound Chinook salmon,

Puget Sound Steelhead, and bull trout, environmental windows (also known as “fish windows”) have

been established for the LDW. These are designated periods (generally from October through February),

when effects of in-water construction on salmon are minimal, largely because juvenile salmon are not

migrating through the waterway during that period. EPA will consult with the National Marine Fisheries

Service and the U. S. Fish and Wildlife Service (Services) to ensure protection of threatened or

endangered species. EPA will prepare a Biological Assessment (BA) for the Services in accordance with

Section 7(c) of ESA; and EPA anticipates that the Services will evaluate the BA and produce a Biological

Opinion (BO) in accordance with Section 7(b) including any reasonable and prudent measures to be

taken, which will guide implementation of the Selected Remedy with respect to the protection of listed

species.

14.3 Cost-Effectiveness

The Selected Remedy is cost-effective and represents a reasonable value for the costs incurred. In making

this determination, the following definition was used: “A remedy shall be cost-effective if its costs are


150

proportional to its overall effectiveness.” (40 CFR 300.430(f)(1)(ii)(D)). EPA evaluated the “overall

effectiveness” of those alternatives that satisfied the threshold criteria (i.e., were both protective of human

health and the environment and ARAR-compliant) by assessing three of the five balancing criteria in

combination (long-term effectiveness and permanence; reduction in toxicity, mobility, and volume

through treatment; and short-term effectiveness). Overall effectiveness was then compared to costs to

determine cost-effectiveness. The relationship of the overall effectiveness of this remedial alternative was

determined to be proportional to its costs and hence this alternative represents a reasonable value for the

money to be spent.

The total estimated capital costs (net present value) to construct the Selected Remedy are $295 million,

and the total estimated operation, maintenance, and monitoring costs (net present value) are

approximately $48 million for a total of $342 million (excluding the cost of source control, which is not

part of the In-waterway Portion of the Site, and excluding the cost of the Early Actions). Less costly

alternatives rely more on technologies such as ENR and MNR that have greater uncertainty as to their

long-term effectiveness. In more costly alternatives, the additional costs are not proportional to the overall

increase in protectiveness.

14.4 Utilization of Permanent Solutions and Alternative Treatment (or Resource Recovery) Technologies to the Maximum Extent Practicable

EPA has determined that the Selected Remedy represents the maximum extent to which permanent

solutions and treatment technologies can be utilized in a practicable manner at the Site. Of those

alternatives that are protective of human health and the environment and comply with ARARs, EPA has

determined that the Selected Remedy provides the best balance of trade-offs in terms of the five balancing

criteria, while also considering the statutory preference for treatment as a principal element and bias

against off-site treatment and disposal and considering State and community acceptance.

14.5 Preference for Treatment as a Principal Element The Selected Remedy does not satisfy the statutory preference for treatment as a principal element of the

remedy because there is no cost-effective means of treating the large quantity of contamination present at

the In-waterway Portion of the Site. The NCP establishes the expectation that treatment will be used to

address the principal threats posed by a site whenever practicable, (40 CFR 300.430[a] [1] [iii] [A]). In

general, principal threat wastes are those source materials considered to be highly toxic or highly mobile

that generally cannot be contained in a reliable manner, or will present a significant risk to human health

or the environment should exposure occur. As discussed in Section 11, EPA has determined that the

contaminated sediments in the LDW outside of the EAAs are not highly mobile or highly toxic. However,

the Selected Remedy does include potential treatment through ENR/in situ treatment using activated

carbon or other sequestering agents if pilot tests are successful. In situ carbon amendment works best at

relatively low levels of contamination.

14.6 Five-Year Review Requirement

Because this remedy will result in hazardous substances, pollutants, or contaminants remaining on-site

above levels that allow for unlimited use and unrestricted exposure, a statutory review will be conducted

within five years after initiation of remedial action to ensure that the remedy is, or will be, protective of

human health and the environment.


151

15 Key Terms

Apparent Effects Threshold (AET) – the concentrations of a specific contaminant in sediment above

which an adverse biological effect always occurs for a particular biological indicator. The lowest AET

(LAET) refers to the most sensitive assay, and a second lowest AET (2LAET) refers to the next most

sensitive assay.

Applicable or Relevant and Appropriate Requirements (ARARs) – ARARs are substantive (as

opposed to administrative) standards, requirements, criteria, or limitations under any federal

environmental law, or promulgated under any state environmental or facility siting law that are more

stringent than under federal law, which must be met or formally waived upon (or before) completion of

final remedial action. See Section 121(d) of CERCLA.

Bed Composition Model (BCM) – along with the Sediment Transport Model, this was a tool used in the

Feasibility Study to predict future contaminant concentrations in LDW sediments during and following

implementation of each of the proposed cleanup alternatives.

CERCLA – the Comprehensive Environmental Response, Compensation, and Liability Act—also known

as Superfund—CERCLA is a federal law which authorizes response actions to reduce the dangers

associated with releases or threats of releases of hazardous substances, pollutants, or contaminants that

may endanger public health or welfare or the environment.

Chronic Daily Intake (CDI) – intake of a contaminant averaged over a lifetime, often adjusted for

absorption efficiency.

Cleanup Screening Level (CSL) – see text box “What are the Sediment Management Standards?”. The

level used to identify cleanup sites and the maximum level for establishing sediment cleanup levels under

the Washington State Sediment Management Standards.

Congener – structurally related chemical substances. For example, there are 202 chemical congeners in

the suite of polychlorinated biphenyls that share the generalized biphenyl organization and differ

according to location of the chlorine.

Contained Aquatic Disposal (CAD) – disposal of dredged sediment in a depression or bermed area at

the bottom of a water body. The area is then capped with clean sediment.

Contaminant of Concern (COC) – a hazardous substance or group of substances that pose unacceptable

risk to human health or the environment.

Enhanced Natural Recovery (ENR) – an active remedial technology which includes placement of a thin

clean sand or sediment layer as a means to accelerate recovery.

Excess Lifetime Cancer Risk – the incremental probability of an individual developing cancer over a

lifetime as a result of exposure to a carcinogen.

Food-Web Model (FWM) – a mathematical model to estimate the relationship among sediment, water,

and tissue contaminant concentrations including higher trophic levels (such as fish).

Hazard Index (HI) – the sum of more than one hazard quotient for multiple substances and/or multiple

exposure pathways. The HI is calculated separately for chronic, subchronic, and shorter-duration

exposures. An HI may be used to evaluate the risk for multiple noncarcinogenic hazardous substances

with similar modes of toxic action.


152

Hazard Quotient (HQ) – a method to summarize the relative level of risk for a single noncarcinogenic

hazardous substance that is based on the ratio of an exposure over a specified time period to a reference

dose.

Human Health Risk Assessment (HHRA) – an assessment to determine potential pathways by which

humans could be exposed to contamination at or from a site. The assessment determines the amount of

exposure, and estimates the resulting level of toxicity.

Imposex- A condition of gender alteration (masculinization) in mollusks associated with exposure to

butyltins.

In situ Treatment – an active remedial technology conducted in place (e.g., without removing sediment).

It includes reactive caps and ENR amendments that enhance breakdown of or bind contaminants.

Institutional controls (ICs) – non-engineered measures that may be selected as remedial or response

actions either by themselves or in combination with engineered remedies, such as administrative and legal

controls that minimize the potential for human exposure to contamination by limiting land or resource

use.

Lowest Observed Adverse Effect Level (LOAEL) – the lowest contaminant concentration documented

to have shown a related negative impact on the reference species either from observation or by

experiment.

Mean Lower Low Water (MLLW) – the average height of the lowest tide recorded at a tide station each

day over the period from 1983 to 2001.

Model Toxics Control Act (MTCA) – a Washington State cleanup law generally similar to CERCLA.

MTCA establishes substantive requirements for cleanup actions (as State ARARs) when those

requirements are more stringent than CERCLA requirements. MTCA includes the SMS and its numerical

and biological criteria for the protection of marine benthic invertebrates.

Monitored Natural Recovery (MNR) – MNR is a passive remedial technology that relies on natural

processes to reduce ecological and human health risks to acceptable levels, while monitoring recovery

over time to verify remedy success.

Natural Background – as defined in MTCA regulations, the concentration of a hazardous substance

consistently present in an environment that has not been influenced by localized human activities. For

some hazardous substances such as PCBs, background conditions may be influenced by global-

distribution patterns.

No Adverse Effects Level (NOAEL) – the level of exposure to an organism at which no biologically or

statistically significant effect is found in exposed test organisms in comparison to a control.

Organic Carbon (OC) Normalized Values –the bioavailability and toxicity of some organic

contaminants in sediments are expected to correlate with the OC in sediment. Accordingly, some SMS

criteria divide the dry weight value by the fraction of OC present in the sample (e.g., 12 mg/kg OC is the

benthic SCO for total PCBs).

Preliminary Remediation Goals (PRGs) – contaminant concentrations that are developed during an

RI/FS and proposed as Cleanup Levels in a Proposed Plan. They are based upon applicable or relevant

and appropriate requirements (ARARs) and other information whenever ARARs are not adequately

protective of all receptors at a site. PRGs may become cleanup levels in the ROD.

http://en.wikipedia.org/wiki/Statistical_significance


153

Principal Threat Waste (PTW) – a source of hazardous substances that is highly toxic or highly mobile,

such as pools of non-aqueous phase liquids, and that generally cannot be reliably contained or would

present a significant risk to human health or the environment should exposure occur.

Propeller Wash, Propeller Scour – (see Vessel Scour)

Reasonable Maximum Exposure (RME) – the risk assessment scenario which portrays the highest level

of human exposure that could reasonably be expected to occur.

Reference Dose (RfD) – an estimate of a daily oral exposure to the human population (including

sensitive subgroups) that is likely to be without an appreciable risk of deleterious effects during a lifetime.

RfDs may be adjusted for uncertainties related to the type of test organism or organ system selected for

the criterion; often this adjustment is by an order of magnitude.

Recovery Categories – categories used in the LDW FS to assign remedial technologies to specific areas

based on information about the potential for sediment contaminant concentrations to be reduced through

natural recovery and/or for subsurface contamination to be exposed due to erosion or scour. As used in

this ROD, Category 1 refers to recovery that is presumed to be limited; Category 2 refers to recovery that

is uncertain, and Category 3 refers to recovery that is predicted to occur with some confidence.

Remedial Action Levels (RALs) – contaminant-specific sediment concentrations designed to identify

specific areas of sediments that require active remediation, taking into consideration the human health and

ecological risk reduction achieved by the different remedial technologies.

Remedial Action Objectives (RAOs) – objectives that describe what the proposed cleanup is expected to

accomplish in order to protect human health and the environment.

Risk-based Threshold Concentrations (RBTCs) – the calculated concentrations in any medium

estimated to be protective of a particular receptor for a given exposure pathway and target risk level.

RBTCs are based on the baseline risk assessments conducted during the RI.

Sediment Cleanup Objective (SCO) – see text box “What are the Sediment Management Standards?”.

SCO represents the level that is the environmental goal for establishing sediment cleanup levels under the

Washington State Sediment Management Standards.

Sediment Transport Model (STM) – a three-dimensional model developed to simulate sediment

movement over a wide range of flow and tidal conditions to inform the type of sediment cleanup

technologies that would be appropriate for the LDW. See also Bed Composition Model.

Slope Factor (SF) - An upper bound, approximating a 95% confidence limit, on the increased cancer risk

from a lifetime exposure to an agent. This estimate, usually expressed in units of proportion (of a

population) affected per mg/kg-day, is generally reserved for use in the low-dose region of the dose-

response relationship, that is, for exposures corresponding to risks less than 1 in 100.

Spatially Weighted Average Concentration (SWAC) – a means of interpreting data across a surface

(such as the LDW), using interpolation methods to generate a mosaic of concentrations which may then

be averaged over the area.

Spud/Spudding – as applied to marine navigation, a spud is a metal pole that is dropped into the

sediment by its own weight and is used for temporary anchorage or stabilization of vessels.

Toxic Equivalent (TEQ) – a single value used to express the joint toxicity of a mixture of compounds

with a similar toxic action, e.g., dioxins/furans or carcinogenic PAHs.

http://en.wikipedia.org/wiki/Order_of_magnitude


154

Tug Scour – (see Vessel Scour)

Upper Confidence Limit (95%) on the Mean (UCL95) – the UCL95 is used to estimate exposure to

human health, fish, and wildlife to concentrations of hazardous substances in the environment. It is

intended to ensure that these concentrations are not underestimated when a number of values are

averaged. Use of this statistic assures no more than a 5% chance that the average of point concentrations

will be exceeded.

Vessel Scour – erosion and movement of sediment due to propeller thrust or due to grounding of vessels.

Vessel scour may cause subsurface contamination to reach the surface.


155

16 References Cited

Arnot, JA and Gobas FAPC . 2004. A food web bioaccumulation model for organic chemicals in aquatic

ecosystems. Environ Toxicol Chem, 23:2343-2355.

Daniell W, Childers J, Gould L, Cummings BJ, Lenhart A. Health Impact Assessment: Proposed Cleanup

Plan for the Lower Duwamish Waterway Superfund Site. Seattle, WA: University of Washington,

Just Health Action, and Duwamish River Cleanup Coalition/Technical Advisory Group. September

2013.

Department of the Navy, 2012. Third Five Year Review Bremerton Naval Complex Bremerton,

Washington. October 2012. http://www.epa.gov/region10/pdf/sites/puget_sound_naval_shipyard/

bnc_3rd_fyr_1012.pdf

Ecology (Washington State Department of Ecology). 1996a. Progress Re-evaluating Puget Sound

Apparent Effects Thresholds (AETs) Volume I: 1994 Amphipod and Echinoderm Larval AETs.

Ecology. 1996b. Sediment Management Standards Cleanup Action Decision: Norfolk CSO. December

1996.

Ecology. 2002. Final Sediment Management Standards Cleanup Action Decision: Duwamish/Diagonal

CSO/SD. July 25, 2002.

Ecology. 2004. Draft Sediment Management Standards Cleanup Action Decision Amendment:

Duwamish/Diagonal CSO/SD. December 14, 2004.

Ecology. 2007. Using Sediment Profile Imaging (SPI) to Evaluate Sediment Quality at Two Cleanup

Sites in Puget Sound Part I - Lower Duwamish Waterway. Publication No. 07-03-025.

Ecology, 2011. Surface Sediment Sampling at Outfalls in the Lower Duwamish Waterway, Seattle, WA.

October 2011.

Ecology 2012. Lower Duwamish Waterway Source Control Strategy. Draft Final.

Ecology, 2013a. Draft Sediment Cleanup Users Manual II: Guidance for Implementing the Cleanup

Provisions of the Sediment Management Standards, Chapter 173-204 WAC. Publication 12-09-057.

December 2013. https://fortress.wa.gov/ecy/publications/publications/1209057.pdf Date accessed: 17

October 2014.Ecology 2013b. Final Cleanup Action Plan Port Gamble Bay, Port Gamble,

Washington. December 2013. On the website of Washington Department of Ecology; Toxics Cleanup

at https://fortress.wa.gov/ecy/gsp/Sitepage.aspx?csid=3444 [Under Site Documents, see “Port Gamble Bay

and Mill Site Final Cleanup Action Plan December 2013.” Date accessed: 17 October 2014.]

Ecology 2013 [multiple]. Cleanup action plans for several sites in the Anacortes Baywide Sites

(Fidalgo/Padilla Bays) or Other Major Puget Sound Cleanup Locations. [The cleanup plans can be

accessed by going to the page for an individual site and looking under headings such as Site

Documents or Documents Available for Public Review. http://www.ecy.wa.gov/programs/tcp/

sites_brochure/psi/anacortes/psi_anacortes_bay.html Date accessed: 17 October 2014.]

EPA (US Environmental Protection Agency). 1988. Guidance for Conducting Remedial Investigations

and Feasibility Studies under CERCLA. EPA/540/G-89/004. U.S. Environmental Protection Agency,

Washington, D.C. 1988.

EPA. 1989. Risk Assessment Guidance for Superfund Volume 1 – Human Health Evaluation Manual

(Part A), EPA/540.—89/002, December 1989 http://www.epa.gov/oswer/riskassessment/ragsa/

http://www.epa.gov/region10/pdf/sites/puget_sound_naval_shipyard/bnc_3rd_fyr_1012.pdf

http://www.epa.gov/region10/pdf/sites/puget_sound_naval_shipyard/bnc_3rd_fyr_1012.pdf

https://fortress.wa.gov/ecy/gsp/Sitepage.aspx?csid=3444

http://www.ecy.wa.gov/programs/tcp/sites_brochure/psi/anacortes/psi_anacortes_bay.html




http://www.epa.gov/oswer/riskassessment/ragsa/


156

EPA 1993. “Guidance for Evaluating Technical Impracticability of Ground-Water Restoration,”

September, 1993. EPA OSWER Directive 9234.2-25

EPA.1995. Land Use in the CERCLA Remedy Selection Process, EPA OSWER Directive 9355.7-04.

May 1995.

EPA 1997. EPA Superfund Record of Decision: Old Navy Dump/Manchester Laboratory

(USEPA/NOAA). http://www.epa.gov/superfund/sites/rods/fulltext/r1097201.pdf Date accessed: 17

October 2014.

EPA. 2000a. A Guide to Developing and Documenting Cost Estimates During the Feasibility Study.

Office of Emergency and Remedial Response, U.S. Environmental Protection Agency, Washington,

D.C. 20460. EPA 540-R-00-002 Office of Solid Waste and Emergency Response (OSWER) 9355.0-

75. July 2000.

EPA 2000b. EPA Superfund Record of Decision: Puget Sound Naval Shipyard Complex, EPA ID:

WA2170023418, OU2, Bremerton, WA. http://www.epa.gov/superfund/sites/rods/fulltext/r1000516.pdf

Date accessed: 17 October 2014.

EPA. 2002a. Role of Background in the CERCLA Cleanup Program, EPA OSWER Directive 9285.6-

07P; April 26, 2002

EPA. 2002b. EPA OSWER directive 9285.6.08, Principles for Managing Contaminated Sediment Risks

at Hazardous Waste Sites. February 12, 2002.

EPA 2002c. Fish Consumption and Environmental Justice: A Report developed from the National

Environmental Justice Advisory Council Meeting of December 3-6, 2001.

EPA. 2004. Risk Assessment Guidance for Superfund (RAGS), Volume I: Human Health Evaluation

Manual (Part E, Supplemental Guidance for Dermal Risk Assessment). EPA/540/R/99/005, OSWER

9285.7-02EP, PB99-963312. Washington, DC. July 2004.

EPA. 2005. Contaminated Sediment Remediation Guidance for Hazardous Waste Sites. Office of Solid

Waste and Emergency Response. U.S. Environmental Protection Agency, Washington, D.C. EPA-

540-R-05-012, OSWER 9355.0-85. December 2005.

http://www.epa.gov/superfund/health/conmedia/sediment/guidance.htm

EPA. 2007. Framework for Selecting and Using Tribal Fish and Shellfish Consumption Rates for Risk-

Based Decision Making at CERCLA and RCRA Cleanup Sites in Puget Sound and the Strait of

Georgia. EPA Region 10 Offices of Environmental Cleanup; Air, Waste and Toxics; and

Environmental Assessment. Seattle, WA. August 2007.

EPA 2009b. Region 10 Superfund, RCRA, LUST, and Brownfields Clean and Green Policy dated August

13, 2009. Seattle, WA.

EPA. 2009c. Third Five Year Review for Commencement Bay/Nearshore Tideflats Superfund Site

Tacoma, Washington. December 23, 2009. http://www.epa.gov/region10/pdf/sites/cb-nt/

cbnt_3rd_5yr_122309.pdf

EPA. 2009d. Five-Year Review Report Pacific Sound Resources Superfund Site Seattle, King County,

Washington. September 2009. http://www.epa.gov/region10/pdf/sites/psr/second_5yr_review_092809.pdf

http://www.epa.gov/superfund/sites/rods/fulltext/r1097201.pdf

http://www.epa.gov/superfund/sites/rods/fulltext/r1000516.pdf

http://www.epa.gov/superfund/health/conmedia/sediment/guidance.htm

http://www.epa.gov/region10/pdf/sites/cb-nt/cbnt_3rd_5yr_122309.pdf

http://www.epa.gov/region10/pdf/sites/cb-nt/cbnt_3rd_5yr_122309.pdf

http://www.epa.gov/region10/pdf/sites/psr/second_5yr_review_092809.pdf


157

EPA. 2012. Five Year Review Report for Wyckoff/Eagle Harbor Superfund Site Kitsap County,

Washington. Third Five-Year Review Report. http://www.epa.gov/region10/pdf/sites/

wyckoff-eagleharbor/wyckoff_3rd_fyr_092712.pdf

EPA. 2013a. Proposed Plan—Lower Duwamish Superfund Site, Appendix B. (Appendix B is the Draft

Environmental Justice Analysis).

EPA 2013b. Statistical Software ProUCL 5.0.00 for Environmental Applications for Data Sets with and

without Nondetect Observations. http://www.epa.gov/osp/hstl/tsc/software.htm#about

EPA 2013c. Record of Decision: Grasse River Superfund Site (a.k.a. Alcoa Aggregation Site), Massena,

St. Lawrence County, New York. http://www.epa.gov/region2/superfund/npl/aluminumcompany/ Date

accessed: 17 October 2014.

EPA 2014a. Environmental Justice Analysis for the Lower Duwamish Waterway Cleanup. October 2014.

EPA 2014b. Lower Duwamish Waterway Environmental Justice Analysis Comments and Responses.

October 2014.

EPA and USACE (US Army Corps of Engineers). 1998. Guidance for In-Situ Subaqueous Capping of

Contaminated Sediments. EPA 905-B96-004, Great Lakes National Program Office, Chicago, IL.

http://www.epa.gov/glnpo/sediment/iscmain/

EPA and Ecology. 2014. Green/Duwamish River Watershed Pollutant Loading Assessment Technical

Approach. Prepared by Tetra Tech, Inc. for the Environmental Protection Agency Region 10 and

Washington State Department of Ecology. September 2014.

Goodrum, P., K. Jenkins, M. Barajas-Albalawi, P. Krause, F. Ellerbusch. 2013. Weighing Biological and

Chemical Lines of Evidence: Sediment Investigation of the Lower Duwamish Waterway.

Presentation. Seventh International Conference on Remediation of Contaminated Sediments, Dallas

TX Feb 4-7, 2013.

Hugla JL, Thome JP. 1999. Effects of polychlorinated biphenyls on liver ultrastructure, hepatic

monooxygenases, and reproductive success in the barbel. Ecotoxicol Environ Saf 42:265-273.

King County. 1999. King County combined sewer overflow water quality assessment for the Duwamish

River and Elliott Bay. Vol 1, Appendix B2, B3, & B4: human health, wildlife, and aquatic life risk

assessments. King County Department of Natural Resources, Seattle, WA.

King County. 2010. Draft BCM Lateral and Input Sensitivity Values for the LDW FS. Memorandum to

AECOM. March 17, 2010. Seattle, WA.

King County. 2014. Analysis of Transport Alternatives for Lower Duwamish Cleanup. Letter with

attachments from Pam Elardo, Department of NaturalResources and Parks - Wastewater Treatment

Division, to Christie True, Director of Depar tment of Natural Resources and Parks. Seattle, WA.

King County, Washington. No date. Webpage: King County Invader: New Zealand Mudsnail.

http://www.kingcounty.gov/environment/animalsAndPlants/biodiversity/threats/Invasives/Mudsnails.aspx

Date accessed: 17 October 2014.

Kondo, Y. 1987. Burrowing depth of infaunal bivalves - observations in living species and its relation to

shell morphology. Trans. Proc. Paleont. Soc. Japan, N. S., No. 148, 306-323.

LDWG. 2008. Sediment Transport Analysis Final Report. Final. Submitted to U.S. EPA, Seattle, WA,

and Ecology, Bellevue, WA.

http://www.epa.gov/region10/pdf/sites/wyckoff-eagleharbor/wyckoff_3rd_fyr_092712.pdf

http://www.epa.gov/region10/pdf/sites/wyckoff-eagleharbor/wyckoff_3rd_fyr_092712.pdf

http://www.epa.gov/osp/hstl/tsc/software.htm#about

http://www.epa.gov/region2/superfund/npl/aluminumcompany/

http://www.epa.gov/glnpo/sediment/iscmain/

http://www.kingcounty.gov/environment/animalsAndPlants/biodiversity/threats/Invasives/Mudsnails.aspx


158

LDWG 2009. Lower Duwamish Waterway Remedial Investigation Report, Appendix B, Baseline Human

Health Risk Assessment. Errata: Adjustment to Tulalip Tribes Seafood Consumption Rates and the

Impact on Risk Estimates. Submitted to U.S. EPA, Seattle, WA, and Ecology, Bellevue, WA.

LDWG. 2010. Lower Duwamish Waterway Remedial Investigation, Remedial Investigation Report.

Final. Submitted to U.S. EPA, Seattle, WA, and Ecology, Bellevue, WA..

LDWG. 2012a. Final Feasibility Study Lower Duwamish Waterway, Seattle, Washington. Submitted to

U.S. EPA, Seattle, WA, and Ecology, Bellevue, WA.

LDWG. 2012b. Technical Memorandum: Supplement to the Feasibility Study for the LDW Superfund

Site, Approaches for Addressing Additional Concerns in Alternative 5C and Development of

Alternative 5C Plus Scenarios. Submitted to U.S. EPA, Seattle, WA, and Ecology, Bellevue, WA.

LDWG. 2013. Development of Final Technology Assignments and Modifications to Alternative 5C Plus

Scenario 5a in Support of EPA’s Selected Remedy. Submitted to U.S. EPA, Seattle, WA, and

Ecology, Bellevue, WA.

LDWG. 2014a. Revised Cost Estimates for EPA’s Proposed Plan Remedy for the Lower Duwamish

Waterway (LDW) Using Updated Present Value Discount Rates. Submitted to U.S. EPA, Seattle,

WA, and Ecology, Bellevue, WA.

LDWG. 2014b. Lower Duwamish Waterway Fishers Study Final Implementation Plan. Prepared by

Windward Environmental LLC for the Environmental Protection Agency Region 10 and Washington

State Department of Ecology. September 2014.

Lotufo et. al. 2014. Lotufo, GR, JL Wilkens, AJ Bednar, JM Biedenbach, BC Suedel, JS Wakeman, MM

Michalsen and G-M Momplaisir. 2014. Regional Applied Research Effort (RARE) Project Task 1b

Bioaccumulation of Arsenic in Mya arenaria Exposed for 60-d to Suspended Sediment and

Undisturbed Bed Sediment from the Lower Duwamish River. EPA Region 10 RARE Program

Report

Magar, V.S., D.B. Chadwick, T.S. Bridges, P.C. Fuchsman, J.M. Conder, T.J. Dekker, J.A. Steevens,

K.E. Gustavson, and M.A. Mills. 2009. Technical Guide: Monitored Natural Recovery at

Contaminated Sediment Sites. Prepared for the Environmental Security Technology Certification

Program. ESTCP Project ER-0622. May 2009.

National Research Council. 2001. A Risk-Management STRATEGY for PCB-Contaminated Sediments.

Committee on Remediation of PCB-Contaminated Sediments. Board on Environmental Studies and

Toxicology Division on Life and Earth Studies. The National Academies Press 500 5th St, NW

Washington D.C., 20001. 2001.

National Research Council. 2007. Sediment Dredging at Superfund Megasites – Assessing the

Effectiveness. Committee on Sediment Dredging at Superfund Megasites. Board on Environmental

Studies and Toxicology, Division of Life and Earth Studies. National Research Council of the

National Academies. The National Academies Press 500 5th St, NW Washington D.C., 20001. 2007.

Oehlmann J., E. Stroben & P. Fioroni, 1991. The morphological expression of imposex in Nucella

lapillus (Linnaeus) Gastropoda: Muricidae). J. moll. Stud. 57: 375–390.

PNNL (Pacific Northwest National Laboratory). No date. Visual Sampling Plan version 6.5.

http://vsp.pnnl.gov/

http://vsp.pnnl.gov/


159

Sechena, Ruth, Connie Nakano, Shiquan Liao, Nayak Polissar, Roseanne Lorenzana, Simon Truong, and

Richard Fenske, 1999. Asian and Pacific Islander Seafood Consumption Study. EPA 910/R-99-003.

Prepared under EPA Environmental Justice Community/University Partnership Grant EQ925003-

001. Seattle, WA. May 27, 1999.

Simpson Tacoma Kraft Company. 1999. St. Paul Waterway Area Remedial Action and Habitat

Restoration Project Final 1998 Monitoring Report. February 1999. http://yosemite.epa.gov/r10/

CLEANUP.NSF/CBNT/St.+Paul+Waterway/$FILE/98Monitoring+Report.pdf

Toy, K.A., N.L. Polissar, S. Liao, and G.D. Mittelstaedt. 1996. A Fish Consumption Survey of the Tulalip

and Squaxin Island Tribes of the Puget Sound Region. Tulalip Tribes, Department of Environment,

Marysville, WA.

USACE. 2008. The Four R’s of Environmental Dredging: Resuspension, Release, Residual, and Risk

Prepared by T.S. Bridges, S. Ells, D. Hayes, D. Mount, S. Nadeau, M.R. Palermo, C. Patmont, and P.

Schroeder. February 2008.

USACE (US Army Corps of Engineers). 2013a. Letter to Allison Hiltner, EPA. USACE Comments on

Lower Duwamish Waterway Proposed Plan. June 13, 1013.

USACE. 2013b. Data Report -Lower Duwamish Waterway, East Waterway, and West Waterway

Subsurface Sediment Characterization. Seattle, Washington. Prepared by HDR Engineering, Inc.,

Science and Engineering for the Environment LLC, and Ken Taylor Associates. URL:

http://www.nws.usace.army.mil/Portals/27/docs/civilworks/dredging/Suitability%20Determinations/2013/

Duwamish_Data_Report_05202013.pdf

USACE — Engineer Research and Development Laboratory (ERDC) Environmental Laboratoy. No

date. New Zealand Mud Snail webpage. http://el.erdc.usace.army.mil/ansrp/ANSIS/html/

potamopyrgus_antipodarum_new_zealand_mud_snail.htm. Date Accessed 17 October 2014.

USACE et al. 2009. OSV Bold Summer 2008 Survey Data Report. (Prepared by USACE Seattle

District Dredged Material Management Program; EPA Region 10; USACE; and Washington

Departments of Natural Resources and Ecology). June 25, 2009.

http://www.nws.usace.army.mil/Portals/27/docs/

civilworks/dredging/Dioxin/Bold%20Data%20Report_Final-red.pdf

WDOH. 2008. Washington Dept. of Health Lower Duwamish Waterway Fact Sheet: DOH 334-139

October 2008.

http://yosemite.epa.gov/r10/CLEANUP.NSF/CBNT/St.+Paul+Waterway/$FILE/98Monitoring+Report.pdf

http://yosemite.epa.gov/r10/CLEANUP.NSF/CBNT/St.+Paul+Waterway/$FILE/98Monitoring+Report.pdf

http://www.nws.usace.army.mil/Portals/27/docs/civilworks/dredging/Suitability%20Determinations/2013/Duwamish_Data_Report_05202013.pdf

http://www.nws.usace.army.mil/Portals/27/docs/civilworks/dredging/Suitability%20Determinations/2013/Duwamish_Data_Report_05202013.pdf

http://el.erdc.usace.army.mil/ansrp/ANSIS/html/potamopyrgus_antipodarum_new_zealand_mud_snail.htm

http://el.erdc.usace.army.mil/ansrp/ANSIS/html/potamopyrgus_antipodarum_new_zealand_mud_snail.htm

http://www.nws.usace.army.mil/Portals/27/docs/civilworks/dredging/Dioxin/Bold%20Data%20Report_Final-red.pdf

http://www.nws.usace.army.mil/Portals/27/docs/civilworks/dredging/Dioxin/Bold%20Data%20Report_Final-red.pdf


160



161

Part 3 Responsiveness Summary [separate volume]


162



163

Attachment 1 Washington State Department of Ecology

Concurrence Letter


164



165

Date post:	04-Jun-2020
Category:	Documents
Upload:	others
View:	1 times
Download:	0 times